fix: optimize HF Hub lookups with smart merging, limited retries, and detailed logging

data/providers.json

@@ -133,7 +133,9 @@
           "type": "embedding",
           "input_price_per_1m": 0.02,
           "output_price_per_1m": 0,
-          "currency": "USD"
+          "currency": "USD",
+          "hf_id": "BAAI/bge-m3",
+          "size_b": 0.57
         }
       ],
       "lastUpdated": "2026-03-02T04:54:46.028Z"
@@ -1036,13 +1038,26 @@
           "hf_id": "meta-llama/Meta-Llama-3-8B-Instruct"
         },
         {
-          "name": "novita/Sao10K/L3-8B-Stheno-v3.2",
+          "name": "alibaba/qwen-turbo",
           "type": "chat",
           "input_price_per_1m": 0.05,
-          "output_price_per_1m": 0.05,
+          "output_price_per_1m": 0.2,
           "currency": "USD",
-          "context_window": 8192,
-          "size_b": 8
+          "context_window": 1000000,
+          "capabilities": [
+            "tools"
+          ]
+        },
+        {
+          "name": "azure/gpt-5-nano",
+          "type": "vision",
+          "input_price_per_1m": 0.05,
+          "output_price_per_1m": 0.4,
+          "currency": "USD",
+          "capabilities": [
+            "reasoning"
+          ],
+          "context_window": 200000
         },
         {
           "name": "novita/meta-llama/llama-3.1-8b-instruct",
@@ -1067,38 +1082,25 @@
           "size_b": 8
         },
         {
-          "name": "openai-responses/gpt-5-nano",
-          "type": "vision",
+          "name": "novita/Sao10K/L3-8B-Stheno-v3.2",
+          "type": "chat",
           "input_price_per_1m": 0.05,
-          "output_price_per_1m": 0.4,
+          "output_price_per_1m": 0.05,
           "currency": "USD",
-          "capabilities": [
-            "vision",
-            "reasoning"
-          ],
-          "context_window": 400000
+          "context_window": 8192,
+          "size_b": 8
         },
         {
-          "name": "azure/gpt-5-nano",
+          "name": "openai-responses/gpt-5-nano",
           "type": "vision",
           "input_price_per_1m": 0.05,
           "output_price_per_1m": 0.4,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 200000
-        },
-        {
-          "name": "alibaba/qwen-turbo",
-          "type": "chat",
-          "input_price_per_1m": 0.05,
-          "output_price_per_1m": 0.2,
-          "currency": "USD",
-          "context_window": 1000000,
-          "capabilities": [
-            "tools"
-          ]
+          "context_window": 400000
         },
         {
           "name": "together/meta-llama/Llama-3.2-3B-Instruct-Turbo",
@@ -1110,21 +1112,21 @@
           "size_b": 3
         },
         {
-          "name": "deepinfra/microsoft/phi-4",
+          "name": "deepinfra/Qwen/Qwen2.5-Coder-32B-Instruct",
           "type": "chat",
           "input_price_per_1m": 0.07,
-          "output_price_per_1m": 0.14,
+          "output_price_per_1m": 0.16,
           "currency": "USD",
-          "context_window": 16384
+          "context_window": 16384,
+          "size_b": 32
         },
         {
-          "name": "deepinfra/Qwen/Qwen2.5-Coder-32B-Instruct",
+          "name": "deepinfra/microsoft/phi-4",
           "type": "chat",
           "input_price_per_1m": 0.07,
-          "output_price_per_1m": 0.16,
+          "output_price_per_1m": 0.14,
           "currency": "USD",
-          "context_window": 16384,
-          "size_b": 32
+          "context_window": 16384
         },
         {
           "name": "novita/gryphe/mythomax-l2-13b",
@@ -1136,16 +1138,18 @@
           "size_b": 13
         },
         {
-          "name": "vertex/gemini-2.5-flash-lite",
-          "type": "vision",
+          "name": "groq/openai/gpt-oss-20b",
+          "type": "chat",
           "input_price_per_1m": 0.1,
-          "output_price_per_1m": 0.4,
+          "output_price_per_1m": 0.5,
           "currency": "USD",
+          "context_window": 131072,
+          "size_b": 20,
           "capabilities": [
-            "vision",
+            "tools",
             "reasoning"
           ],
-          "context_window": 1048576
+          "hf_id": "openai/gpt-oss-20b"
         },
         {
           "name": "openai/gpt-4.1-nano",
@@ -1159,27 +1163,30 @@
           "context_window": 1047576
         },
         {
-          "name": "mistral/devstral-small-latest",
-          "type": "chat",
+          "name": "google/gemini-2.0-flash-001",
+          "type": "vision",
           "input_price_per_1m": 0.1,
-          "output_price_per_1m": 0.3,
+          "output_price_per_1m": 0.4,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "tools"
+            "vision"
           ],
-          "hf_id": "mistralai/Devstral-Small-2507"
+          "context_window": 1048576
         },
         {
-          "name": "mistral/mistral-small-2503",
-          "type": "chat",
+          "name": "google/gemini-2.5-flash-lite",
+          "type": "vision",
           "input_price_per_1m": 0.1,
-          "output_price_per_1m": 0.3,
+          "output_price_per_1m": 0.4,
           "currency": "USD",
-          "context_window": 32768
+          "capabilities": [
+            "vision",
+            "reasoning"
+          ],
+          "context_window": 1048576
         },
         {
-          "name": "mistral/devstral-small-2507",
+          "name": "mistral/devstral-small-latest",
           "type": "chat",
           "input_price_per_1m": 0.1,
           "output_price_per_1m": 0.3,
@@ -1191,41 +1198,27 @@
           "hf_id": "mistralai/Devstral-Small-2507"
         },
         {
-          "name": "openai-responses/gpt-4.1-nano",
-          "type": "vision",
-          "input_price_per_1m": 0.1,
-          "output_price_per_1m": 0.4,
-          "currency": "USD",
-          "capabilities": [
-            "vision"
-          ],
-          "context_window": 1047576
-        },
-        {
-          "name": "together/meta-llama/Meta-Llama-3-8B-Instruct-Lite",
+          "name": "mistral/mistral-small-2503",
           "type": "chat",
           "input_price_per_1m": 0.1,
-          "output_price_per_1m": 0.1,
+          "output_price_per_1m": 0.3,
           "currency": "USD",
-          "context_window": 8192,
-          "size_b": 8
+          "context_window": 32768
         },
         {
-          "name": "deepinfra/Qwen/Qwen3-32B",
+          "name": "mistral/devstral-small-2507",
           "type": "chat",
           "input_price_per_1m": 0.1,
           "output_price_per_1m": 0.3,
           "currency": "USD",
-          "context_window": 40000,
-          "size_b": 32,
+          "context_window": 131072,
           "capabilities": [
-            "tools",
-            "reasoning"
+            "tools"
           ],
-          "hf_id": "Qwen/Qwen3-32B"
+          "hf_id": "mistralai/Devstral-Small-2507"
         },
         {
-          "name": "google/gemini-2.5-flash-lite",
+          "name": "vertex/gemini-2.5-flash-lite",
           "type": "vision",
           "input_price_per_1m": 0.1,
           "output_price_per_1m": 0.4,
@@ -1237,7 +1230,7 @@
           "context_window": 1048576
         },
         {
-          "name": "google/gemini-2.0-flash-001",
+          "name": "azure/openai-responses/gpt-4.1-nano",
           "type": "vision",
           "input_price_per_1m": 0.1,
           "output_price_per_1m": 0.4,
@@ -1245,10 +1238,10 @@
           "capabilities": [
             "vision"
           ],
-          "context_window": 1048576
+          "context_window": 1047576
         },
         {
-          "name": "azure/openai-responses/gpt-4.1-nano",
+          "name": "azure/gpt-4.1-nano",
           "type": "vision",
           "input_price_per_1m": 0.1,
           "output_price_per_1m": 0.4,
@@ -1259,7 +1252,16 @@
           "context_window": 1047576
         },
         {
-          "name": "azure/gpt-4.1-nano",
+          "name": "together/meta-llama/Meta-Llama-3-8B-Instruct-Lite",
+          "type": "chat",
+          "input_price_per_1m": 0.1,
+          "output_price_per_1m": 0.1,
+          "currency": "USD",
+          "context_window": 8192,
+          "size_b": 8
+        },
+        {
+          "name": "openai-responses/gpt-4.1-nano",
           "type": "vision",
           "input_price_per_1m": 0.1,
           "output_price_per_1m": 0.4,
@@ -1270,18 +1272,18 @@
           "context_window": 1047576
         },
         {
-          "name": "groq/openai/gpt-oss-20b",
+          "name": "deepinfra/Qwen/Qwen3-32B",
           "type": "chat",
           "input_price_per_1m": 0.1,
-          "output_price_per_1m": 0.5,
+          "output_price_per_1m": 0.3,
           "currency": "USD",
-          "context_window": 131072,
-          "size_b": 20,
+          "context_window": 40000,
+          "size_b": 32,
           "capabilities": [
             "tools",
             "reasoning"
           ],
-          "hf_id": "openai/gpt-oss-20b"
+          "hf_id": "Qwen/Qwen3-32B"
         },
         {
           "name": "deepinfra/meta-llama/Llama-3.3-70B-Instruct-Turbo",
@@ -1331,18 +1333,18 @@
           "size_b": 8
         },
         {
-          "name": "novita/deepseek/deepseek-r1-distill-qwen-14b",
+          "name": "groq/openai/gpt-oss-120b",
           "type": "chat",
           "input_price_per_1m": 0.15,
-          "output_price_per_1m": 0.15,
+          "output_price_per_1m": 0.75,
           "currency": "USD",
-          "context_window": 128000,
-          "size_b": 14,
+          "context_window": 131072,
+          "size_b": 120,
           "capabilities": [
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1"
+          "hf_id": "openai/gpt-oss-120b"
         },
         {
           "name": "openai/gpt-4o-mini",
@@ -1356,16 +1358,26 @@
           "context_window": 128000
         },
         {
-          "name": "mistral/mistral-small-latest",
-          "type": "vision",
+          "name": "nebius/openai/gpt-oss-120b",
+          "type": "chat",
           "input_price_per_1m": 0.15,
           "output_price_per_1m": 0.6,
           "currency": "USD",
           "capabilities": [
-            "vision",
             "reasoning"
           ],
-          "context_window": 256000
+          "context_window": 131000,
+          "size_b": 120,
+          "hf_id": "openai/gpt-oss-120b"
+        },
+        {
+          "name": "parasail/parasail-qwen3-235b-a22b-instruct-2507",
+          "type": "chat",
+          "input_price_per_1m": 0.15,
+          "output_price_per_1m": 0.85,
+          "currency": "USD",
+          "context_window": 262144,
+          "size_b": 235
         },
         {
           "name": "mistral/mistral-small-2603",
@@ -1382,40 +1394,30 @@
           "hf_id": "mistralai/Mistral-Small-4-119B-2603"
         },
         {
-          "name": "parasail/parasail-qwen3-235b-a22b-instruct-2507",
-          "type": "chat",
-          "input_price_per_1m": 0.15,
-          "output_price_per_1m": 0.85,
-          "currency": "USD",
-          "context_window": 262144,
-          "size_b": 235
-        },
-        {
-          "name": "nebius/openai/gpt-oss-120b",
-          "type": "chat",
+          "name": "mistral/mistral-small-latest",
+          "type": "vision",
           "input_price_per_1m": 0.15,
           "output_price_per_1m": 0.6,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 131000,
-          "size_b": 120,
-          "hf_id": "openai/gpt-oss-120b"
+          "context_window": 256000
         },
         {
-          "name": "groq/openai/gpt-oss-120b",
+          "name": "novita/deepseek/deepseek-r1-distill-qwen-14b",
           "type": "chat",
           "input_price_per_1m": 0.15,
-          "output_price_per_1m": 0.75,
+          "output_price_per_1m": 0.15,
           "currency": "USD",
-          "context_window": 131072,
-          "size_b": 120,
+          "context_window": 128000,
+          "size_b": 14,
           "capabilities": [
             "tools",
             "reasoning"
           ],
-          "hf_id": "openai/gpt-oss-120b"
+          "hf_id": "deepseek-ai/DeepSeek-R1"
         },
         {
           "name": "novita/mistralai/mistral-nemo",
@@ -1439,18 +1441,29 @@
           "size_b": 8
         },
         {
-          "name": "novita/meta-llama/llama-4-maverick-17b-128e-instruct-fp8",
+          "name": "openai/gpt-5.4-nano",
           "type": "vision",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 0.85,
+          "output_price_per_1m": 1.25,
           "currency": "USD",
-          "context_window": 1048576,
-          "size_b": 17,
           "capabilities": [
             "vision",
-            "tools"
+            "reasoning"
           ],
-          "hf_id": "meta-llama/Llama-4-Maverick-17B-128E-Instruct"
+          "context_window": 400000
+        },
+        {
+          "name": "alibaba/qwen3-30b-a3b-instruct-2507",
+          "type": "vision",
+          "input_price_per_1m": 0.2,
+          "output_price_per_1m": 0.8,
+          "currency": "USD",
+          "capabilities": [
+            "vision"
+          ],
+          "context_window": 131072,
+          "size_b": 30,
+          "hf_id": "Qwen/Qwen3-30B-A3B-Instruct-2507"
         },
         {
           "name": "novita/qwen/qwen3-235b-a22b-fp8",
@@ -1467,44 +1480,34 @@
           "hf_id": "Qwen/Qwen3-235B-A22B"
         },
         {
-          "name": "openai/gpt-5.4-nano",
+          "name": "novita/meta-llama/llama-4-maverick-17b-128e-instruct-fp8",
           "type": "vision",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 1.25,
+          "output_price_per_1m": 0.85,
           "currency": "USD",
+          "context_window": 1048576,
+          "size_b": 17,
           "capabilities": [
             "vision",
-            "reasoning"
+            "tools"
           ],
-          "context_window": 400000
+          "hf_id": "meta-llama/Llama-4-Maverick-17B-128E-Instruct"
         },
         {
-          "name": "openai-responses/gpt-5.4-nano",
+          "name": "xai/grok-4-fast-non-reasoning",
           "type": "vision",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 1.25,
+          "output_price_per_1m": 0.5,
           "currency": "USD",
+          "context_window": 2000000,
           "capabilities": [
             "vision",
-            "reasoning"
-          ],
-          "context_window": 400000
-        },
-        {
-          "name": "xai/grok-4-fast-non-reasoning",
-          "type": "vision",
-          "input_price_per_1m": 0.2,
-          "output_price_per_1m": 0.5,
-          "currency": "USD",
-          "context_window": 2000000,
-          "capabilities": [
-            "vision",
-            "tools",
+            "tools",
             "reasoning"
           ]
         },
         {
-          "name": "xai/grok-4-1-fast-non-reasoning",
+          "name": "xai/grok-4-1-fast-reasoning",
           "type": "vision",
           "input_price_per_1m": 0.2,
           "output_price_per_1m": 0.5,
@@ -1515,15 +1518,15 @@
           "context_window": 2000000
         },
         {
-          "name": "xai/grok-code-fast-1",
+          "name": "xai/grok-4-1-fast-non-reasoning",
           "type": "vision",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 1.5,
+          "output_price_per_1m": 0.5,
           "currency": "USD",
           "capabilities": [
             "vision"
           ],
-          "context_window": 256000
+          "context_window": 2000000
         },
         {
           "name": "xai/grok-4-fast",
@@ -1539,15 +1542,15 @@
           ]
         },
         {
-          "name": "xai/grok-4-1-fast-reasoning",
+          "name": "xai/grok-code-fast-1",
           "type": "vision",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 0.5,
+          "output_price_per_1m": 1.5,
           "currency": "USD",
           "capabilities": [
             "vision"
           ],
-          "context_window": 2000000
+          "context_window": 256000
         },
         {
           "name": "together/meta-llama/LlamaGuard-2-8b",
@@ -1559,18 +1562,16 @@
           "size_b": 8
         },
         {
-          "name": "deepinfra/zai-org/GLM-4.5-Air",
-          "type": "chat",
+          "name": "openai-responses/gpt-5.4-nano",
+          "type": "vision",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 1.1,
+          "output_price_per_1m": 1.25,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "tools",
+            "vision",
             "reasoning"
           ],
-          "hf_id": "zai-org/GLM-4.5-Air",
-          "size_b": 110.5
+          "context_window": 400000
         },
         {
           "name": "deepinfra/Qwen/Qwen3-235B-A22B",
@@ -1587,17 +1588,31 @@
           "hf_id": "Qwen/Qwen3-235B-A22B"
         },
         {
-          "name": "alibaba/qwen3-30b-a3b-instruct-2507",
-          "type": "vision",
+          "name": "deepinfra/zai-org/GLM-4.5-Air",
+          "type": "chat",
           "input_price_per_1m": 0.2,
-          "output_price_per_1m": 0.8,
+          "output_price_per_1m": 1.1,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
-            "vision"
+            "tools",
+            "reasoning"
           ],
+          "hf_id": "zai-org/GLM-4.5-Air",
+          "size_b": 110.5
+        },
+        {
+          "name": "deepinfra/meta-llama/Llama-3.3-70B-Instruct",
+          "type": "chat",
+          "input_price_per_1m": 0.23,
+          "output_price_per_1m": 0.4,
+          "currency": "USD",
           "context_window": 131072,
-          "size_b": 30,
-          "hf_id": "Qwen/Qwen3-30B-A3B-Instruct-2507"
+          "size_b": 70,
+          "capabilities": [
+            "tools"
+          ],
+          "hf_id": "meta-llama/Llama-3.3-70B-Instruct"
         },
         {
           "name": "deepinfra/deepseek-ai/DeepSeek-R1-Distill-Llama-70B",
@@ -1635,20 +1650,7 @@
           "hf_id": "Qwen/Qwen2.5-VL-72B-Instruct"
         },
         {
-          "name": "deepinfra/meta-llama/Llama-3.3-70B-Instruct",
-          "type": "chat",
-          "input_price_per_1m": 0.23,
-          "output_price_per_1m": 0.4,
-          "currency": "USD",
-          "context_window": 131072,
-          "size_b": 70,
-          "capabilities": [
-            "tools"
-          ],
-          "hf_id": "meta-llama/Llama-3.3-70B-Instruct"
-        },
-        {
-          "name": "vertex/gemini-3.1-flash-lite-preview",
+          "name": "google/gemini-3.1-flash-lite-preview",
           "type": "vision",
           "input_price_per_1m": 0.25,
           "output_price_per_1m": 1.5,
@@ -1668,38 +1670,38 @@
           "size_b": 7
         },
         {
-          "name": "openai-responses/gpt-5-mini",
+          "name": "vertex/gemini-3.1-flash-lite-preview",
           "type": "vision",
           "input_price_per_1m": 0.25,
-          "output_price_per_1m": 2,
+          "output_price_per_1m": 1.5,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
-          "context_window": 400000
+          "context_window": 1048576
         },
         {
-          "name": "google/gemini-3.1-flash-lite-preview",
+          "name": "azure/gpt-5-mini",
           "type": "vision",
           "input_price_per_1m": 0.25,
-          "output_price_per_1m": 1.5,
+          "output_price_per_1m": 2,
           "currency": "USD",
           "capabilities": [
-            "vision"
+            "reasoning"
           ],
-          "context_window": 1048576
+          "context_window": 200000
         },
         {
-          "name": "azure/gpt-5-mini",
+          "name": "openai-responses/gpt-5-mini",
           "type": "vision",
           "input_price_per_1m": 0.25,
           "output_price_per_1m": 2,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 400000
         },
         {
           "name": "novita/deepseek/deepseek-v3.2",
@@ -1711,7 +1713,8 @@
             "reasoning"
           ],
           "context_window": 163840,
-          "hf_id": "deepseek-ai/DeepSeek-V3.2"
+          "hf_id": "deepseek-ai/DeepSeek-V3.2",
+          "size_b": 685.4
         },
         {
           "name": "deepseek/deepseek-chat",
@@ -1734,46 +1737,53 @@
           "context_window": 128000
         },
         {
-          "name": "novita/minimax-m2.7",
-          "type": "vision",
+          "name": "nebius/minimax/minimax-m2.1",
+          "type": "chat",
           "input_price_per_1m": 0.3,
           "output_price_per_1m": 1.2,
           "currency": "USD",
           "capabilities": [
-            "vision",
             "reasoning"
           ],
           "context_window": 200000,
-          "hf_id": "MiniMaxAI/MiniMax-M2.7"
+          "hf_id": "MiniMaxAI/MiniMax-M2.1",
+          "size_b": 228.7
         },
         {
-          "name": "novita/deepseek/deepseek-r1-distill-qwen-32b",
+          "name": "nebius/deepseek-ai/DeepSeek-V3.2",
           "type": "chat",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 0.3,
+          "output_price_per_1m": 0.45,
           "currency": "USD",
-          "context_window": 12800,
-          "size_b": 32,
           "capabilities": [
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B"
+          "context_window": 164000,
+          "hf_id": "deepseek-ai/DeepSeek-V3.2",
+          "size_b": 685.4
         },
         {
-          "name": "novita/MiniMax-M2.7",
+          "name": "parasail/parasail-gemma3-27b-it",
+          "type": "chat",
+          "input_price_per_1m": 0.3,
+          "output_price_per_1m": 0.5,
+          "currency": "USD",
+          "context_window": 128000,
+          "size_b": 27
+        },
+        {
+          "name": "alibaba/qwen3-coder-flash",
           "type": "vision",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 1.2,
+          "output_price_per_1m": 1.5,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
-          "context_window": 200000,
-          "hf_id": "MiniMaxAI/MiniMax-M2.7"
+          "context_window": 1048576
         },
         {
-          "name": "vertex/gemini-2.5-flash",
+          "name": "google/gemini-2.5-flash",
           "type": "vision",
           "input_price_per_1m": 0.3,
           "output_price_per_1m": 2.5,
@@ -1785,7 +1795,15 @@
           "context_window": 1048576
         },
         {
-          "name": "vertex/gemini-2.5-flash-image",
+          "name": "mistral/codestral-latest",
+          "type": "chat",
+          "input_price_per_1m": 0.3,
+          "output_price_per_1m": 0.9,
+          "currency": "USD",
+          "context_window": 131072
+        },
+        {
+          "name": "vertex/gemini-2.5-flash",
           "type": "vision",
           "input_price_per_1m": 0.3,
           "output_price_per_1m": 2.5,
@@ -1797,7 +1815,7 @@
           "context_window": 1048576
         },
         {
-          "name": "coding/gemini-2.5-flash",
+          "name": "vertex/gemini-2.5-flash-image",
           "type": "vision",
           "input_price_per_1m": 0.3,
           "output_price_per_1m": 2.5,
@@ -1809,90 +1827,71 @@
           "context_window": 1048576
         },
         {
-          "name": "mistral/codestral-latest",
-          "type": "chat",
-          "input_price_per_1m": 0.3,
-          "output_price_per_1m": 0.9,
-          "currency": "USD",
-          "context_window": 131072
-        },
-        {
-          "name": "xai/grok-3-mini",
-          "type": "chat",
+          "name": "novita/minimax-m2.7",
+          "type": "vision",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 0.5,
+          "output_price_per_1m": 1.2,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 131072
+          "context_window": 200000,
+          "hf_id": "MiniMaxAI/MiniMax-M2.7"
         },
         {
-          "name": "together/Qwen/Qwen2.5-7B-Instruct-Turbo",
+          "name": "novita/deepseek/deepseek-r1-distill-qwen-32b",
           "type": "chat",
           "input_price_per_1m": 0.3,
           "output_price_per_1m": 0.3,
           "currency": "USD",
-          "context_window": 32768,
-          "size_b": 7
-        },
-        {
-          "name": "deepinfra/deepseek-ai/DeepSeek-V3.1",
-          "type": "chat",
-          "input_price_per_1m": 0.3,
-          "output_price_per_1m": 1,
-          "currency": "USD",
-          "context_window": 163840,
+          "context_window": 12800,
+          "size_b": 32,
           "capabilities": [
-            "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-V3.1-Terminus"
-        },
-        {
-          "name": "parasail/parasail-gemma3-27b-it",
-          "type": "chat",
-          "input_price_per_1m": 0.3,
-          "output_price_per_1m": 0.5,
-          "currency": "USD",
-          "context_window": 128000,
-          "size_b": 27
+          "hf_id": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B"
         },
         {
-          "name": "google/gemini-2.5-flash",
+          "name": "novita/MiniMax-M2.7",
           "type": "vision",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 2.5,
+          "output_price_per_1m": 1.2,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 1048576
+          "context_window": 200000,
+          "hf_id": "MiniMaxAI/MiniMax-M2.7"
         },
         {
-          "name": "nebius/deepseek-ai/DeepSeek-V3.2",
+          "name": "minimaxi/MiniMax-M2",
           "type": "chat",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 0.45,
+          "output_price_per_1m": 1.2,
           "currency": "USD",
+          "context_window": 200000,
           "capabilities": [
+            "tools",
             "reasoning"
           ],
-          "context_window": 164000,
-          "hf_id": "deepseek-ai/DeepSeek-V3.2"
+          "hf_id": "MiniMaxAI/MiniMax-M2",
+          "size_b": 228.7
         },
         {
-          "name": "nebius/minimax/minimax-m2.1",
-          "type": "chat",
+          "name": "minimaxi/MiniMax-M2.5",
+          "type": "vision",
           "input_price_per_1m": 0.3,
           "output_price_per_1m": 1.2,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
           "context_window": 200000,
-          "hf_id": "MiniMaxAI/MiniMax-M2.1"
+          "hf_id": "MiniMaxAI/MiniMax-M2.5",
+          "size_b": 228.7
         },
         {
           "name": "minimaxi/MiniMax-M2.7",
@@ -1908,41 +1907,49 @@
           "hf_id": "MiniMaxAI/MiniMax-M2.7"
         },
         {
-          "name": "minimaxi/MiniMax-M2",
+          "name": "xai/grok-3-mini",
           "type": "chat",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 1.2,
+          "output_price_per_1m": 0.5,
           "currency": "USD",
-          "context_window": 200000,
           "capabilities": [
-            "tools",
             "reasoning"
           ],
-          "hf_id": "MiniMaxAI/MiniMax-M2"
+          "context_window": 131072
         },
         {
-          "name": "minimaxi/MiniMax-M2.5",
+          "name": "together/Qwen/Qwen2.5-7B-Instruct-Turbo",
+          "type": "chat",
+          "input_price_per_1m": 0.3,
+          "output_price_per_1m": 0.3,
+          "currency": "USD",
+          "context_window": 32768,
+          "size_b": 7
+        },
+        {
+          "name": "coding/gemini-2.5-flash",
           "type": "vision",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 1.2,
+          "output_price_per_1m": 2.5,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 200000,
-          "hf_id": "MiniMaxAI/MiniMax-M2.5"
+          "context_window": 1048576
         },
         {
-          "name": "alibaba/qwen3-coder-flash",
-          "type": "vision",
+          "name": "deepinfra/deepseek-ai/DeepSeek-V3.1",
+          "type": "chat",
           "input_price_per_1m": 0.3,
-          "output_price_per_1m": 1.5,
+          "output_price_per_1m": 1,
           "currency": "USD",
+          "context_window": 163840,
           "capabilities": [
-            "vision"
+            "tools",
+            "reasoning"
           ],
-          "context_window": 1048576
+          "hf_id": "deepseek-ai/DeepSeek-V3.1-Terminus"
         },
         {
           "name": "deepinfra/meta-llama/Llama-3.2-90B-Vision-Instruct",
@@ -1980,51 +1987,47 @@
           "hf_id": "meta-llama/Llama-3.3-70B-Instruct"
         },
         {
-          "name": "novita/deepseek/deepseek-v3-0324",
-          "type": "chat",
+          "name": "openai/gpt-4.1-mini",
+          "type": "vision",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.3,
+          "output_price_per_1m": 1.6,
           "currency": "USD",
-          "context_window": 128000,
           "capabilities": [
-            "tools",
-            "reasoning"
+            "vision"
           ],
-          "size_b": 685,
-          "hf_id": "deepseek-ai/DeepSeek-V3-0324"
-        },
-        {
-          "name": "novita/deepseek/deepseek-v3-turbo",
-          "type": "chat",
-          "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.3,
-          "currency": "USD",
-          "context_window": 128000
+          "context_window": 1047576
         },
         {
-          "name": "openai/gpt-4.1-mini",
-          "type": "vision",
+          "name": "nebius/Qwen/Qwen3-Coder-480B-A35B-Instruct",
+          "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.6,
+          "output_price_per_1m": 1.8,
           "currency": "USD",
+          "context_window": 262000,
+          "size_b": 480,
           "capabilities": [
-            "vision"
+            "tools"
           ],
-          "context_window": 1047576
+          "hf_id": "Qwen/Qwen3-Coder-480B-A35B-Instruct"
         },
         {
-          "name": "mistral/devstral-latest",
+          "name": "nebius/zai-org/GLM-4.7",
           "type": "chat",
           "input_price_per_1m": 0.4,
           "output_price_per_1m": 2,
           "currency": "USD",
-          "context_window": 256000
+          "capabilities": [
+            "reasoning"
+          ],
+          "context_window": 200000,
+          "hf_id": "zai-org/GLM-4.7",
+          "size_b": 358.3
         },
         {
-          "name": "mistral/devstral-medium-2507",
+          "name": "alibaba/qwen-plus",
           "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 2,
+          "output_price_per_1m": 1.2,
           "currency": "USD",
           "context_window": 131072,
           "capabilities": [
@@ -2040,28 +2043,23 @@
           "context_window": 131072
         },
         {
-          "name": "openai-responses/gpt-4.1-mini",
-          "type": "vision",
+          "name": "mistral/devstral-medium-2507",
+          "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.6,
+          "output_price_per_1m": 2,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
-            "vision"
-          ],
-          "context_window": 1047576
+            "tools"
+          ]
         },
         {
-          "name": "deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct",
+          "name": "mistral/devstral-latest",
           "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.6,
+          "output_price_per_1m": 2,
           "currency": "USD",
-          "context_window": 262144,
-          "size_b": 480,
-          "capabilities": [
-            "tools"
-          ],
-          "hf_id": "Qwen/Qwen3-Coder-480B-A35B-Instruct"
+          "context_window": 256000
         },
         {
           "name": "azure/gpt-4.1-mini",
@@ -2086,75 +2084,91 @@
           "context_window": 1047576
         },
         {
-          "name": "nebius/zai-org/GLM-4.7",
+          "name": "novita/deepseek/deepseek-v3-0324",
           "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 2,
+          "output_price_per_1m": 1.3,
           "currency": "USD",
+          "context_window": 128000,
           "capabilities": [
+            "tools",
             "reasoning"
           ],
-          "context_window": 200000,
-          "hf_id": "zai-org/GLM-4.7"
+          "size_b": 685,
+          "hf_id": "deepseek-ai/DeepSeek-V3-0324"
         },
         {
-          "name": "nebius/Qwen/Qwen3-Coder-480B-A35B-Instruct",
+          "name": "novita/deepseek/deepseek-v3-turbo",
           "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.8,
+          "output_price_per_1m": 1.3,
+          "currency": "USD",
+          "context_window": 128000
+        },
+        {
+          "name": "openai-responses/gpt-4.1-mini",
+          "type": "vision",
+          "input_price_per_1m": 0.4,
+          "output_price_per_1m": 1.6,
           "currency": "USD",
-          "context_window": 262000,
-          "size_b": 480,
           "capabilities": [
-            "tools"
+            "vision"
           ],
-          "hf_id": "Qwen/Qwen3-Coder-480B-A35B-Instruct"
+          "context_window": 1047576
         },
         {
-          "name": "alibaba/qwen-plus",
+          "name": "deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct",
           "type": "chat",
           "input_price_per_1m": 0.4,
-          "output_price_per_1m": 1.2,
+          "output_price_per_1m": 1.6,
           "currency": "USD",
-          "context_window": 131072,
+          "context_window": 262144,
+          "size_b": 480,
           "capabilities": [
             "tools"
-          ]
+          ],
+          "hf_id": "Qwen/Qwen3-Coder-480B-A35B-Instruct"
         },
         {
-          "name": "vertex/gemini-3.1-flash-image-preview",
-          "type": "vision",
+          "name": "nebius/deepseek-ai/DeepSeek-V3-0324",
+          "type": "chat",
           "input_price_per_1m": 0.5,
-          "output_price_per_1m": 2,
+          "output_price_per_1m": 1.5,
           "currency": "USD",
+          "context_window": 128000,
           "capabilities": [
-            "vision",
+            "tools",
             "reasoning"
           ],
-          "context_window": 131072
+          "size_b": 685,
+          "hf_id": "deepseek-ai/DeepSeek-V3-0324"
         },
         {
-          "name": "vertex/gemini-3-flash-preview",
-          "type": "vision",
+          "name": "nebius/moonshotai/Kimi-K2-Instruct",
+          "type": "chat",
           "input_price_per_1m": 0.5,
-          "output_price_per_1m": 3,
+          "output_price_per_1m": 2.4,
           "currency": "USD",
+          "context_window": 131000,
           "capabilities": [
-            "vision",
-            "reasoning"
+            "tools"
           ],
-          "context_window": 1048576
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
-          "name": "mistral/mistral-large-latest",
-          "type": "chat",
+          "name": "nebius/moonshotai/kimi-k2.5",
+          "type": "vision",
           "input_price_per_1m": 0.5,
-          "output_price_per_1m": 1.5,
+          "output_price_per_1m": 2.5,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "tools"
-          ]
+            "vision",
+            "reasoning"
+          ],
+          "context_window": 256000,
+          "hf_id": "moonshotai/Kimi-K2.5",
+          "size_b": 1058.6
         },
         {
           "name": "google/gemini-3.1-flash-image-preview",
@@ -2181,44 +2195,39 @@
           "context_window": 1048576
         },
         {
-          "name": "nebius/moonshotai/Kimi-K2-Instruct",
+          "name": "mistral/mistral-large-latest",
           "type": "chat",
           "input_price_per_1m": 0.5,
-          "output_price_per_1m": 2.4,
+          "output_price_per_1m": 1.5,
           "currency": "USD",
-          "context_window": 131000,
+          "context_window": 131072,
           "capabilities": [
             "tools"
-          ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          ]
         },
         {
-          "name": "nebius/deepseek-ai/DeepSeek-V3-0324",
-          "type": "chat",
+          "name": "vertex/gemini-3-flash-preview",
+          "type": "vision",
           "input_price_per_1m": 0.5,
-          "output_price_per_1m": 1.5,
+          "output_price_per_1m": 3,
           "currency": "USD",
-          "context_window": 128000,
           "capabilities": [
-            "tools",
+            "vision",
             "reasoning"
           ],
-          "size_b": 685,
-          "hf_id": "deepseek-ai/DeepSeek-V3-0324"
+          "context_window": 1048576
         },
         {
-          "name": "nebius/moonshotai/kimi-k2.5",
+          "name": "vertex/gemini-3.1-flash-image-preview",
           "type": "vision",
           "input_price_per_1m": 0.5,
-          "output_price_per_1m": 2.5,
+          "output_price_per_1m": 2,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 256000,
-          "hf_id": "moonshotai/Kimi-K2.5",
-          "size_b": 1058.6
+          "context_window": 131072
         },
         {
           "name": "novita/meta-llama/llama-3-70b-instruct",
@@ -2255,7 +2264,8 @@
             "reasoning"
           ],
           "context_window": 163840,
-          "hf_id": "deepseek-ai/DeepSeek-V3.2"
+          "hf_id": "deepseek-ai/DeepSeek-V3.2",
+          "size_b": 685.4
         },
         {
           "name": "novita/moonshotai/kimi-k2-instruct",
@@ -2267,36 +2277,37 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
-          "name": "novita/zai-org/glm-4.6",
+          "name": "zai/GLM-4.7",
           "type": "chat",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.2,
           "currency": "USD",
-          "context_window": 204800,
           "capabilities": [
-            "tools",
             "reasoning"
           ],
-          "hf_id": "zai-org/GLM-4.6"
+          "context_window": 200000,
+          "hf_id": "zai-org/GLM-4.7",
+          "size_b": 358.3
         },
         {
-          "name": "novita/zai-org/glm-4.5",
+          "name": "zai/GLM-4.5",
           "type": "chat",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.2,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "tools",
             "reasoning"
           ],
-          "hf_id": "zai-org/GLM-4.5"
+          "context_window": 131072,
+          "hf_id": "zai-org/GLM-4.5",
+          "size_b": 358.3
         },
         {
-          "name": "zai/GLM-4.5",
+          "name": "zai/GLM-4.6",
           "type": "chat",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.2,
@@ -2304,32 +2315,37 @@
           "capabilities": [
             "reasoning"
           ],
-          "context_window": 131072,
-          "hf_id": "zai-org/GLM-4.5"
+          "context_window": 200000,
+          "hf_id": "zai-org/GLM-4.6",
+          "size_b": 356.8
         },
         {
-          "name": "zai/GLM-4.6",
+          "name": "nebius/zai-org/GLM-4.5",
           "type": "chat",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.2,
           "currency": "USD",
+          "context_window": 128000,
           "capabilities": [
+            "tools",
             "reasoning"
           ],
-          "context_window": 200000,
-          "hf_id": "zai-org/GLM-4.6"
+          "hf_id": "zai-org/GLM-4.5",
+          "size_b": 358.3
         },
         {
-          "name": "zai/GLM-4.7",
-          "type": "chat",
+          "name": "alibaba/qwen3.5",
+          "type": "vision",
           "input_price_per_1m": 0.6,
-          "output_price_per_1m": 2.2,
+          "output_price_per_1m": 3.6,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 200000,
-          "hf_id": "zai-org/GLM-4.7"
+          "context_window": 256000,
+          "size_b": 9,
+          "hf_id": "Qwen/Qwen3.5-9B"
         },
         {
           "name": "moonshot/kimi-k2-0711-preview",
@@ -2341,7 +2357,8 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
           "name": "moonshot/kimi-k2.5",
@@ -2368,7 +2385,8 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Thinking"
+          "hf_id": "moonshotai/Kimi-K2-Thinking",
+          "size_b": 1058.1
         },
         {
           "name": "moonshot/kimi-k2-0905-preview",
@@ -2380,7 +2398,8 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct-0905"
+          "hf_id": "moonshotai/Kimi-K2-Instruct-0905",
+          "size_b": 1026.5
         },
         {
           "name": "moonshot/kimi-k2-thinking",
@@ -2392,10 +2411,11 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
-          "name": "deepinfra/zai-org/GLM-4.5",
+          "name": "novita/zai-org/glm-4.5",
           "type": "chat",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.2,
@@ -2405,23 +2425,25 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "zai-org/GLM-4.5"
+          "hf_id": "zai-org/GLM-4.5",
+          "size_b": 358.3
         },
         {
-          "name": "nebius/zai-org/GLM-4.5",
+          "name": "novita/zai-org/glm-4.6",
           "type": "chat",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.2,
           "currency": "USD",
-          "context_window": 128000,
+          "context_window": 204800,
           "capabilities": [
             "tools",
             "reasoning"
           ],
-          "hf_id": "zai-org/GLM-4.5"
+          "hf_id": "zai-org/GLM-4.6",
+          "size_b": 356.8
         },
         {
-          "name": "minimaxi/MiniMax-M2.7-highspeed",
+          "name": "minimaxi/MiniMax-M2.5-highspeed",
           "type": "vision",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.4,
@@ -2431,10 +2453,11 @@
             "reasoning"
           ],
           "context_window": 200000,
-          "hf_id": "MiniMaxAI/MiniMax-M2.7"
+          "hf_id": "MiniMaxAI/MiniMax-M2.5",
+          "size_b": 228.7
         },
         {
-          "name": "minimaxi/MiniMax-M2.5-highspeed",
+          "name": "minimaxi/MiniMax-M2.7-highspeed",
           "type": "vision",
           "input_price_per_1m": 0.6,
           "output_price_per_1m": 2.4,
@@ -2444,21 +2467,21 @@
             "reasoning"
           ],
           "context_window": 200000,
-          "hf_id": "MiniMaxAI/MiniMax-M2.5"
+          "hf_id": "MiniMaxAI/MiniMax-M2.7"
         },
         {
-          "name": "alibaba/qwen3.5",
-          "type": "vision",
+          "name": "deepinfra/zai-org/GLM-4.5",
+          "type": "chat",
           "input_price_per_1m": 0.6,
-          "output_price_per_1m": 3.6,
+          "output_price_per_1m": 2.2,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
-            "vision",
+            "tools",
             "reasoning"
           ],
-          "context_window": 256000,
-          "size_b": 9,
-          "hf_id": "Qwen/Qwen3.5-9B"
+          "hf_id": "zai-org/GLM-4.5",
+          "size_b": 358.3
         },
         {
           "name": "novita/microsoft/wizardlm-2-8x22b",
@@ -2482,13 +2505,16 @@
           "context_window": 400000
         },
         {
-          "name": "novita/deepseek/deepseek-prover-v2-671b",
-          "type": "chat",
+          "name": "parasail/parasail-qwen25-vl-72b-instruct",
+          "type": "vision",
           "input_price_per_1m": 0.7,
-          "output_price_per_1m": 2.5,
+          "output_price_per_1m": 0.7,
           "currency": "USD",
-          "context_window": 160000,
-          "size_b": 671
+          "capabilities": [
+            "vision"
+          ],
+          "context_window": 32768,
+          "size_b": 72
         },
         {
           "name": "novita/deepseek/deepseek-r1-turbo",
@@ -2501,19 +2527,17 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1"
+          "hf_id": "deepseek-ai/DeepSeek-R1",
+          "size_b": 684.5
         },
         {
-          "name": "parasail/parasail-qwen25-vl-72b-instruct",
-          "type": "vision",
+          "name": "novita/deepseek/deepseek-prover-v2-671b",
+          "type": "chat",
           "input_price_per_1m": 0.7,
-          "output_price_per_1m": 0.7,
+          "output_price_per_1m": 2.5,
           "currency": "USD",
-          "capabilities": [
-            "vision"
-          ],
-          "context_window": 32768,
-          "size_b": 72
+          "context_window": 160000,
+          "size_b": 671
         },
         {
           "name": "openai/gpt-5.4-mini",
@@ -2537,7 +2561,21 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 400000
+        },
+        {
+          "name": "nebius/deepseek-ai/DeepSeek-R1-0528",
+          "type": "chat",
+          "input_price_per_1m": 0.8,
+          "output_price_per_1m": 2.4,
+          "currency": "USD",
+          "context_window": 164000,
+          "capabilities": [
+            "tools",
+            "reasoning"
+          ],
+          "hf_id": "deepseek-ai/DeepSeek-R1-0528",
+          "size_b": 684.5
         },
         {
           "name": "novita/deepseek/deepseek-r1-distill-llama-70b",
@@ -2575,17 +2613,17 @@
           "size_b": 405
         },
         {
-          "name": "nebius/deepseek-ai/DeepSeek-R1-0528",
+          "name": "deepinfra/deepseek-ai/DeepSeek-V3",
           "type": "chat",
-          "input_price_per_1m": 0.8,
-          "output_price_per_1m": 2.4,
+          "input_price_per_1m": 0.85,
+          "output_price_per_1m": 0.9,
           "currency": "USD",
-          "context_window": 164000,
+          "context_window": 128000,
           "capabilities": [
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1-0528"
+          "hf_id": "deepseek-ai/DeepSeek-V3.1-Terminus"
         },
         {
           "name": "deepinfra/deepseek-ai/DeepSeek-R1",
@@ -2598,20 +2636,8 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1"
-        },
-        {
-          "name": "deepinfra/deepseek-ai/DeepSeek-V3",
-          "type": "chat",
-          "input_price_per_1m": 0.85,
-          "output_price_per_1m": 0.9,
-          "currency": "USD",
-          "context_window": 128000,
-          "capabilities": [
-            "tools",
-            "reasoning"
-          ],
-          "hf_id": "deepseek-ai/DeepSeek-V3.1-Terminus"
+          "hf_id": "deepseek-ai/DeepSeek-R1",
+          "size_b": 684.5
         },
         {
           "name": "alibaba/qwen3-max",
@@ -2625,26 +2651,26 @@
           "context_window": 262144
         },
         {
-          "name": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo",
+          "name": "together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
           "type": "chat",
           "input_price_per_1m": 0.88,
           "output_price_per_1m": 0.88,
           "currency": "USD",
           "context_window": 131072,
-          "size_b": 70,
-          "capabilities": [
-            "tools"
-          ],
-          "hf_id": "meta-llama/Llama-3.3-70B-Instruct"
+          "size_b": 70
         },
         {
-          "name": "together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
+          "name": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo",
           "type": "chat",
           "input_price_per_1m": 0.88,
           "output_price_per_1m": 0.88,
           "currency": "USD",
           "context_window": 131072,
-          "size_b": 70
+          "size_b": 70,
+          "capabilities": [
+            "tools"
+          ],
+          "hf_id": "meta-llama/Llama-3.3-70B-Instruct"
         },
         {
           "name": "novita/deepseek/deepseek_v3",
@@ -2668,41 +2694,47 @@
           "context_window": 131072
         },
         {
-          "name": "perplexity/sonar",
+          "name": "zai/GLM-5",
           "type": "vision",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 1,
+          "output_price_per_1m": 3.2,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "vision"
-          ]
+            "vision",
+            "reasoning"
+          ],
+          "context_window": 200000,
+          "hf_id": "zai-org/GLM-5",
+          "size_b": 753.9
         },
         {
-          "name": "bedrock/claude-haiku-4-5",
-          "type": "vision",
+          "name": "groq/moonshotai/kimi-k2-instruct",
+          "type": "chat",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 5,
+          "output_price_per_1m": 3,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
-            "vision"
+            "tools"
           ],
-          "context_window": 200000
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
-          "name": "novita/GLM-5",
+          "name": "groq/moonshotai/Kimi-K2-Instruct-0905",
           "type": "chat",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 3.2,
+          "output_price_per_1m": 3,
           "currency": "USD",
+          "context_window": 256000,
           "capabilities": [
-            "reasoning"
+            "tools"
           ],
-          "context_window": 202800,
-          "hf_id": "zai-org/GLM-5"
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
-          "name": "vertex/claude-haiku-4-5",
+          "name": "anthropic/claude-haiku-4-5",
           "type": "vision",
           "input_price_per_1m": 1,
           "output_price_per_1m": 5,
@@ -2725,20 +2757,18 @@
           "context_window": 200000
         },
         {
-          "name": "zai/GLM-5",
+          "name": "alibaba/qwen3-coder-plus",
           "type": "vision",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 3.2,
+          "output_price_per_1m": 5,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
-          "context_window": 200000,
-          "hf_id": "zai-org/GLM-5"
+          "context_window": 1048576
         },
         {
-          "name": "anthropic/claude-haiku-4-5",
+          "name": "bedrock/claude-haiku-4-5",
           "type": "vision",
           "input_price_per_1m": 1,
           "output_price_per_1m": 5,
@@ -2749,39 +2779,39 @@
           "context_window": 200000
         },
         {
-          "name": "alibaba/qwen3-coder-plus",
+          "name": "perplexity/sonar",
           "type": "vision",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 5,
+          "output_price_per_1m": 1,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
             "vision"
-          ],
-          "context_window": 1048576
+          ]
         },
         {
-          "name": "groq/moonshotai/Kimi-K2-Instruct-0905",
-          "type": "chat",
+          "name": "vertex/claude-haiku-4-5",
+          "type": "vision",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 3,
+          "output_price_per_1m": 5,
           "currency": "USD",
-          "context_window": 256000,
           "capabilities": [
-            "tools"
+            "vision"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "context_window": 200000
         },
         {
-          "name": "groq/moonshotai/kimi-k2-instruct",
+          "name": "novita/GLM-5",
           "type": "chat",
           "input_price_per_1m": 1,
-          "output_price_per_1m": 3,
+          "output_price_per_1m": 3.2,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "tools"
+            "reasoning"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "context_window": 202800,
+          "hf_id": "zai-org/GLM-5",
+          "size_b": 753.9
         },
         {
           "name": "openai/o3-mini",
@@ -2795,7 +2825,7 @@
           "context_window": 200000
         },
         {
-          "name": "openai-responses/o4-mini",
+          "name": "azure/o4-mini",
           "type": "vision",
           "input_price_per_1m": 1.1,
           "output_price_per_1m": 4.4,
@@ -2817,7 +2847,7 @@
           "context_window": 200000
         },
         {
-          "name": "azure/o4-mini",
+          "name": "openai-responses/o4-mini",
           "type": "vision",
           "input_price_per_1m": 1.1,
           "output_price_per_1m": 4.4,
@@ -2837,7 +2867,8 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
           "name": "together/Qwen/Qwen2.5-72B-Instruct-Turbo",
@@ -2849,7 +2880,7 @@
           "size_b": 72
         },
         {
-          "name": "vertex/gemini-2.5-pro",
+          "name": "openai/gpt-5.1",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
@@ -2858,10 +2889,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 1048576
+          "context_window": 400000
         },
         {
-          "name": "openai/gpt-5-chat",
+          "name": "openai/gpt-5.1-chat",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
@@ -2873,7 +2904,7 @@
           "context_window": 128000
         },
         {
-          "name": "openai/gpt-5.1",
+          "name": "openai/gpt-5-chat",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
@@ -2882,10 +2913,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 128000
         },
         {
-          "name": "openai/gpt-5.1-chat",
+          "name": "google/gemini-2.5-pro",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
@@ -2894,10 +2925,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 128000
+          "context_window": 1048576
         },
         {
-          "name": "coding/gemini-2.5-pro",
+          "name": "vertex/gemini-2.5-pro",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
@@ -2909,40 +2940,39 @@
           "context_window": 1048576
         },
         {
-          "name": "openai-responses/gpt-5",
+          "name": "azure/gpt-5.1",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
           "currency": "USD",
           "capabilities": [
-            "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 200000
         },
         {
-          "name": "openai-responses/gpt-5.1",
+          "name": "azure/gpt-5",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
           "currency": "USD",
           "capabilities": [
-            "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 200000
         },
         {
-          "name": "openai-responses/gpt-5.1-codex",
-          "type": "vision",
+          "name": "together/deepseek-ai/DeepSeek-V3",
+          "type": "chat",
           "input_price_per_1m": 1.25,
-          "output_price_per_1m": 10,
+          "output_price_per_1m": 1.25,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
-            "vision",
+            "tools",
             "reasoning"
           ],
-          "context_window": 400000
+          "hf_id": "deepseek-ai/DeepSeek-V3.1-Terminus"
         },
         {
           "name": "openai-responses/gpt-5-codex",
@@ -2957,20 +2987,19 @@
           "context_window": 400000
         },
         {
-          "name": "together/deepseek-ai/DeepSeek-V3",
-          "type": "chat",
+          "name": "openai-responses/gpt-5.1",
+          "type": "vision",
           "input_price_per_1m": 1.25,
-          "output_price_per_1m": 1.25,
+          "output_price_per_1m": 10,
           "currency": "USD",
-          "context_window": 131072,
           "capabilities": [
-            "tools",
+            "vision",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-V3.1-Terminus"
+          "context_window": 400000
         },
         {
-          "name": "google/gemini-2.5-pro",
+          "name": "openai-responses/gpt-5",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
@@ -2979,29 +3008,31 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 1048576
+          "context_window": 400000
         },
         {
-          "name": "azure/gpt-5",
+          "name": "openai-responses/gpt-5.1-codex",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 400000
         },
         {
-          "name": "azure/gpt-5.1",
+          "name": "coding/gemini-2.5-pro",
           "type": "vision",
           "input_price_per_1m": 1.25,
           "output_price_per_1m": 10,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 1048576
         },
         {
           "name": "novita/sao10k/l31-70b-euryale-v2.2",
@@ -3033,7 +3064,7 @@
           ]
         },
         {
-          "name": "openai/gpt-5.3-chat",
+          "name": "openai/gpt-5.2",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3042,10 +3073,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 128000
+          "context_window": 400000
         },
         {
-          "name": "openai/gpt-5.2",
+          "name": "openai/gpt-5.2-chat",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3054,10 +3085,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 128000
         },
         {
-          "name": "openai/gpt-5.2-chat",
+          "name": "openai/gpt-5.3-chat",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3069,19 +3100,18 @@
           "context_window": 128000
         },
         {
-          "name": "openai-responses/gpt-5.2-codex",
+          "name": "azure/gpt-5.2",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
           "currency": "USD",
           "capabilities": [
-            "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 200000
         },
         {
-          "name": "openai-responses/gpt-5.3-codex",
+          "name": "azure/openai-responses/gpt-5.2-codex",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3093,7 +3123,7 @@
           "context_window": 400000
         },
         {
-          "name": "openai-responses/gpt-5.3-chat",
+          "name": "azure/gpt-5.3-codex",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3102,10 +3132,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 128000
+          "context_window": 400000
         },
         {
-          "name": "openai-responses/gpt-5.2",
+          "name": "openai-responses/gpt-5.3-codex",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3117,7 +3147,7 @@
           "context_window": 400000
         },
         {
-          "name": "azure/gpt-5.3-codex",
+          "name": "openai-responses/gpt-5.3-chat",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3126,10 +3156,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 128000
         },
         {
-          "name": "azure/openai-responses/gpt-5.2-codex",
+          "name": "openai-responses/gpt-5.2-codex",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
@@ -3141,86 +3171,71 @@
           "context_window": 400000
         },
         {
-          "name": "azure/gpt-5.2",
+          "name": "openai-responses/gpt-5.2",
           "type": "vision",
           "input_price_per_1m": 1.75,
           "output_price_per_1m": 14,
           "currency": "USD",
           "capabilities": [
+            "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 400000
         },
         {
-          "name": "perplexity/sonar-reasoning-pro",
+          "name": "openai/o4-mini-deep-research",
           "type": "vision",
           "input_price_per_1m": 2,
           "output_price_per_1m": 8,
           "currency": "USD",
-          "context_window": 131072,
-          "capabilities": [
-            "vision",
-            "reasoning"
-          ]
-        },
-        {
-          "name": "vertex/gemini-3.1-pro-preview",
-          "type": "vision",
-          "input_price_per_1m": 2,
-          "output_price_per_1m": 12,
-          "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 1048576
+          "context_window": 100000
         },
         {
-          "name": "vertex/gemini-3-pro-preview",
+          "name": "openai/gpt-4.1",
           "type": "vision",
           "input_price_per_1m": 2,
-          "output_price_per_1m": 12,
+          "output_price_per_1m": 8,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
-          "context_window": 1048576
+          "context_window": 1047576
         },
         {
-          "name": "vertex/gemini-3-pro-image-preview",
-          "type": "vision",
+          "name": "nebius/deepseek-ai/DeepSeek-V3-0324-fast",
+          "type": "chat",
           "input_price_per_1m": 2,
-          "output_price_per_1m": 12,
+          "output_price_per_1m": 6,
           "currency": "USD",
-          "capabilities": [
-            "vision",
-            "reasoning"
-          ],
-          "context_window": 1048576
+          "context_window": 128000
         },
         {
-          "name": "openai/gpt-4.1",
+          "name": "google/gemini-3.1-pro-preview",
           "type": "vision",
           "input_price_per_1m": 2,
-          "output_price_per_1m": 8,
+          "output_price_per_1m": 12,
           "currency": "USD",
           "capabilities": [
-            "vision"
+            "vision",
+            "reasoning"
           ],
-          "context_window": 1047576
+          "context_window": 1048576
         },
         {
-          "name": "openai/o4-mini-deep-research",
+          "name": "google/gemini-3-pro-preview",
           "type": "vision",
           "input_price_per_1m": 2,
-          "output_price_per_1m": 8,
+          "output_price_per_1m": 12,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 100000
+          "context_window": 1048576
         },
         {
           "name": "mistral/pixtral-large-latest",
@@ -3231,38 +3246,31 @@
           "context_window": 131072
         },
         {
-          "name": "openai-responses/gpt-4.1",
+          "name": "perplexity/sonar-reasoning-pro",
           "type": "vision",
           "input_price_per_1m": 2,
           "output_price_per_1m": 8,
           "currency": "USD",
+          "context_window": 131072,
           "capabilities": [
-            "vision"
-          ],
-          "context_window": 1047576
-        },
-        {
-          "name": "xai/grok-2-1212",
-          "type": "chat",
-          "input_price_per_1m": 2,
-          "output_price_per_1m": 10,
-          "currency": "USD",
-          "context_window": 131072
+            "vision",
+            "reasoning"
+          ]
         },
         {
-          "name": "xai/grok-4.2-beta",
+          "name": "vertex/gemini-3.1-pro-preview",
           "type": "vision",
           "input_price_per_1m": 2,
-          "output_price_per_1m": 6,
+          "output_price_per_1m": 12,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 2000000
+          "context_window": 1048576
         },
         {
-          "name": "google/gemini-3-pro-preview",
+          "name": "vertex/gemini-3-pro-preview",
           "type": "vision",
           "input_price_per_1m": 2,
           "output_price_per_1m": 12,
@@ -3274,7 +3282,7 @@
           "context_window": 1048576
         },
         {
-          "name": "google/gemini-3.1-pro-preview",
+          "name": "vertex/gemini-3-pro-image-preview",
           "type": "vision",
           "input_price_per_1m": 2,
           "output_price_per_1m": 12,
@@ -3308,26 +3316,38 @@
           "context_window": 1047576
         },
         {
-          "name": "nebius/deepseek-ai/DeepSeek-V3-0324-fast",
+          "name": "xai/grok-2-1212",
           "type": "chat",
           "input_price_per_1m": 2,
+          "output_price_per_1m": 10,
+          "currency": "USD",
+          "context_window": 131072
+        },
+        {
+          "name": "xai/grok-4.2-beta",
+          "type": "vision",
+          "input_price_per_1m": 2,
           "output_price_per_1m": 6,
           "currency": "USD",
-          "context_window": 128000
+          "capabilities": [
+            "vision",
+            "reasoning"
+          ],
+          "context_window": 2000000
         },
         {
-          "name": "openai/gpt-4o-2024-11-20",
+          "name": "openai-responses/gpt-4.1",
           "type": "vision",
-          "input_price_per_1m": 2.5,
-          "output_price_per_1m": 10,
+          "input_price_per_1m": 2,
+          "output_price_per_1m": 8,
           "currency": "USD",
           "capabilities": [
             "vision"
           ],
-          "context_window": 128000
+          "context_window": 1047576
         },
         {
-          "name": "openai/gpt-4o-2024-08-06",
+          "name": "openai/gpt-4o",
           "type": "vision",
           "input_price_per_1m": 2.5,
           "output_price_per_1m": 10,
@@ -3338,7 +3358,7 @@
           "context_window": 128000
         },
         {
-          "name": "openai/gpt-4o-2024-05-13",
+          "name": "openai/gpt-4o-2024-08-06",
           "type": "vision",
           "input_price_per_1m": 2.5,
           "output_price_per_1m": 10,
@@ -3349,19 +3369,18 @@
           "context_window": 128000
         },
         {
-          "name": "openai/gpt-5.4",
+          "name": "openai/gpt-4o-2024-05-13",
           "type": "vision",
           "input_price_per_1m": 2.5,
-          "output_price_per_1m": 15,
+          "output_price_per_1m": 10,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
-          "context_window": 1050000
+          "context_window": 128000
         },
         {
-          "name": "openai/gpt-4o",
+          "name": "openai/gpt-4o-2024-11-20",
           "type": "vision",
           "input_price_per_1m": 2.5,
           "output_price_per_1m": 10,
@@ -3372,7 +3391,7 @@
           "context_window": 128000
         },
         {
-          "name": "openai-responses/gpt-5.4",
+          "name": "openai/gpt-5.4",
           "type": "vision",
           "input_price_per_1m": 2.5,
           "output_price_per_1m": 15,
@@ -3396,18 +3415,19 @@
           "context_window": 1050000
         },
         {
-          "name": "perplexity/sonar-pro",
+          "name": "openai-responses/gpt-5.4",
           "type": "vision",
-          "input_price_per_1m": 3,
+          "input_price_per_1m": 2.5,
           "output_price_per_1m": 15,
           "currency": "USD",
-          "context_window": 204800,
           "capabilities": [
-            "vision"
-          ]
+            "vision",
+            "reasoning"
+          ],
+          "context_window": 1050000
         },
         {
-          "name": "bedrock/claude-sonnet-4-5",
+          "name": "anthropic/claude-sonnet-4-6",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3416,10 +3436,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 1000000
         },
         {
-          "name": "bedrock/claude-sonnet-4-6",
+          "name": "anthropic/claude-sonnet-4-5",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3431,7 +3451,7 @@
           "context_window": 1000000
         },
         {
-          "name": "bedrock/claude-sonnet-4",
+          "name": "anthropic/claude-sonnet-4",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3440,10 +3460,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 1000000
         },
         {
-          "name": "bedrock/claude-3-7-sonnet",
+          "name": "bedrock/claude-sonnet-4-6",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3452,10 +3472,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 1000000
         },
         {
-          "name": "vertex/claude-sonnet-4-5",
+          "name": "bedrock/claude-sonnet-4-5",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3467,7 +3487,7 @@
           "context_window": 200000
         },
         {
-          "name": "vertex/claude-sonnet-4",
+          "name": "bedrock/claude-3-7-sonnet",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3479,7 +3499,7 @@
           "context_window": 200000
         },
         {
-          "name": "vertex/claude-3-7-sonnet",
+          "name": "bedrock/claude-sonnet-4",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3491,30 +3511,29 @@
           "context_window": 200000
         },
         {
-          "name": "vertex/claude-3-5-sonnet",
+          "name": "perplexity/sonar-pro",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
           "currency": "USD",
+          "context_window": 204800,
           "capabilities": [
             "vision"
-          ],
-          "context_window": 200000
+          ]
         },
         {
-          "name": "coding/claude-sonnet-4-20250514",
+          "name": "vertex/claude-3-5-sonnet",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
           "context_window": 200000
         },
         {
-          "name": "anthropic/claude-sonnet-4-5",
+          "name": "vertex/claude-3-7-sonnet",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3523,10 +3542,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 1000000
+          "context_window": 200000
         },
         {
-          "name": "anthropic/claude-sonnet-4-6",
+          "name": "vertex/claude-sonnet-4-5",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3535,10 +3554,10 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 1000000
+          "context_window": 200000
         },
         {
-          "name": "anthropic/claude-sonnet-4",
+          "name": "vertex/claude-sonnet-4",
           "type": "vision",
           "input_price_per_1m": 3,
           "output_price_per_1m": 15,
@@ -3547,7 +3566,7 @@
             "vision",
             "reasoning"
           ],
-          "context_window": 1000000
+          "context_window": 200000
         },
         {
           "name": "xai/grok-4",
@@ -3571,7 +3590,20 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1"
+          "hf_id": "deepseek-ai/DeepSeek-R1",
+          "size_b": 684.5
+        },
+        {
+          "name": "coding/claude-sonnet-4-20250514",
+          "type": "vision",
+          "input_price_per_1m": 3,
+          "output_price_per_1m": 15,
+          "currency": "USD",
+          "capabilities": [
+            "vision",
+            "reasoning"
+          ],
+          "context_window": 200000
         },
         {
           "name": "novita/deepseek/deepseek-r1",
@@ -3584,10 +3616,11 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1"
+          "hf_id": "deepseek-ai/DeepSeek-R1",
+          "size_b": 684.5
         },
         {
-          "name": "bedrock/claude-opus-4-5",
+          "name": "anthropic/claude-opus-4-5",
           "type": "vision",
           "input_price_per_1m": 5,
           "output_price_per_1m": 25,
@@ -3599,7 +3632,7 @@
           "context_window": 200000
         },
         {
-          "name": "bedrock/claude-opus-4-6",
+          "name": "anthropic/claude-opus-4-6",
           "type": "vision",
           "input_price_per_1m": 5,
           "output_price_per_1m": 25,
@@ -3611,19 +3644,18 @@
           "context_window": 1000000
         },
         {
-          "name": "vertex/claude-opus-4-6",
+          "name": "openai/chatgpt-4o",
           "type": "vision",
           "input_price_per_1m": 5,
-          "output_price_per_1m": 25,
+          "output_price_per_1m": 15,
           "currency": "USD",
           "capabilities": [
-            "vision",
-            "reasoning"
+            "vision"
           ],
-          "context_window": 1000000
+          "context_window": 128000
         },
         {
-          "name": "vertex/claude-opus-4-5",
+          "name": "bedrock/claude-opus-4-5",
           "type": "vision",
           "input_price_per_1m": 5,
           "output_price_per_1m": 25,
@@ -3635,18 +3667,19 @@
           "context_window": 200000
         },
         {
-          "name": "openai/chatgpt-4o",
+          "name": "bedrock/claude-opus-4-6",
           "type": "vision",
           "input_price_per_1m": 5,
-          "output_price_per_1m": 15,
+          "output_price_per_1m": 25,
           "currency": "USD",
           "capabilities": [
-            "vision"
+            "vision",
+            "reasoning"
           ],
-          "context_window": 128000
+          "context_window": 1000000
         },
         {
-          "name": "anthropic/claude-opus-4-6",
+          "name": "vertex/claude-opus-4-6",
           "type": "vision",
           "input_price_per_1m": 5,
           "output_price_per_1m": 25,
@@ -3658,7 +3691,7 @@
           "context_window": 1000000
         },
         {
-          "name": "anthropic/claude-opus-4-5",
+          "name": "vertex/claude-opus-4-5",
           "type": "vision",
           "input_price_per_1m": 5,
           "output_price_per_1m": 25,
@@ -3693,7 +3726,7 @@
           "context_window": 100000
         },
         {
-          "name": "bedrock/claude-opus-4",
+          "name": "anthropic/claude-opus-4-1",
           "type": "vision",
           "input_price_per_1m": 15,
           "output_price_per_1m": 75,
@@ -3705,7 +3738,7 @@
           "context_window": 200000
         },
         {
-          "name": "vertex/claude-opus-4",
+          "name": "anthropic/claude-opus-4",
           "type": "vision",
           "input_price_per_1m": 15,
           "output_price_per_1m": 75,
@@ -3717,10 +3750,10 @@
           "context_window": 200000
         },
         {
-          "name": "vertex/claude-opus-4-1",
+          "name": "openai/o1",
           "type": "vision",
           "input_price_per_1m": 15,
-          "output_price_per_1m": 75,
+          "output_price_per_1m": 60,
           "currency": "USD",
           "capabilities": [
             "vision",
@@ -3729,10 +3762,10 @@
           "context_window": 200000
         },
         {
-          "name": "openai/o1",
+          "name": "bedrock/claude-opus-4",
           "type": "vision",
           "input_price_per_1m": 15,
-          "output_price_per_1m": 60,
+          "output_price_per_1m": 75,
           "currency": "USD",
           "capabilities": [
             "vision",
@@ -3741,7 +3774,7 @@
           "context_window": 200000
         },
         {
-          "name": "coding/claude-opus-4-20250514",
+          "name": "vertex/claude-opus-4-1",
           "type": "vision",
           "input_price_per_1m": 15,
           "output_price_per_1m": 75,
@@ -3753,7 +3786,7 @@
           "context_window": 200000
         },
         {
-          "name": "anthropic/claude-opus-4-1",
+          "name": "vertex/claude-opus-4",
           "type": "vision",
           "input_price_per_1m": 15,
           "output_price_per_1m": 75,
@@ -3765,28 +3798,28 @@
           "context_window": 200000
         },
         {
-          "name": "anthropic/claude-opus-4",
+          "name": "openai-responses/gpt-5-pro",
           "type": "vision",
           "input_price_per_1m": 15,
-          "output_price_per_1m": 75,
+          "output_price_per_1m": 120,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 200000
+          "context_window": 400000
         },
         {
-          "name": "openai-responses/gpt-5-pro",
+          "name": "coding/claude-opus-4-20250514",
           "type": "vision",
           "input_price_per_1m": 15,
-          "output_price_per_1m": 120,
+          "output_price_per_1m": 75,
           "currency": "USD",
           "capabilities": [
             "vision",
             "reasoning"
           ],
-          "context_window": 400000
+          "context_window": 200000
         },
         {
           "name": "openai-responses/o3-pro",
@@ -3813,7 +3846,7 @@
           "context_window": 1050000
         },
         {
-          "name": "openai-responses/gpt-5.4-pro",
+          "name": "azure/openai-responses/gpt-5.4-pro",
           "type": "vision",
           "input_price_per_1m": 30,
           "output_price_per_1m": 180,
@@ -3825,7 +3858,7 @@
           "context_window": 1050000
         },
         {
-          "name": "azure/openai-responses/gpt-5.4-pro",
+          "name": "openai-responses/gpt-5.4-pro",
           "type": "vision",
           "input_price_per_1m": 30,
           "output_price_per_1m": 180,
@@ -3897,7 +3930,8 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct"
+          "hf_id": "moonshotai/Kimi-K2-Instruct",
+          "size_b": 1026.5
         },
         {
           "name": "Qwen3-Coder-480B-A35B-Instruct (base)",
@@ -4093,7 +4127,8 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "zai-org/GLM-4.5"
+          "hf_id": "zai-org/GLM-4.5",
+          "size_b": 358.3
         },
         {
           "name": "GLM-4.5-Air (base)",
@@ -4120,7 +4155,8 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1-0528"
+          "hf_id": "deepseek-ai/DeepSeek-R1-0528",
+          "size_b": 684.5
         },
         {
           "name": "DeepSeek-R1-0528 (base)",
@@ -4133,7 +4169,8 @@
             "tools",
             "reasoning"
           ],
-          "hf_id": "deepseek-ai/DeepSeek-R1-0528"
+          "hf_id": "deepseek-ai/DeepSeek-R1-0528",
+          "size_b": 684.5
         },
         {
           "name": "DeepSeek-V3-0324 (fast)",
@@ -4492,7 +4529,8 @@
           "type": "audio",
           "input_price_per_1m": 0.003,
           "output_price_per_1m": 0,
-          "currency": "EUR"
+          "currency": "EUR",
+          "hf_id": "openai/whisper-large-v3"
         },
         {
           "name": "holo2-30b-a3b",
@@ -4950,7 +4988,8 @@
           "input_price_per_1m": 0.017,
           "output_price_per_1m": 0.11,
           "currency": "USD",
-          "hf_id": "ibm-granite/granite-4.0-h-micro"
+          "hf_id": "ibm-granite/granite-4.0-h-micro",
+          "size_b": 3.2
         },
         {
           "name": "google/gemma-3n-e4b-it",
@@ -5277,7 +5316,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 31.2
         },
         {
           "name": "qwen/qwen3-14b",
@@ -5527,7 +5567,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 309.8
         },
         {
           "name": "alibaba/tongyi-deepresearch-30b-a3b",
@@ -5617,7 +5658,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 199.4
         },
         {
           "name": "mistralai/mistral-small-creative",
@@ -5841,7 +5883,8 @@
           "hf_id": "Qwen/Qwen3-Coder-Next",
           "capabilities": [
             "tools"
-          ]
+          ],
+          "size_b": 79.7
         },
         {
           "name": "allenai/olmo-3-7b-think",
@@ -6229,7 +6272,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 228.7
         },
         {
           "name": "allenai/molmo-2-8b",
@@ -6342,7 +6386,8 @@
           "hf_id": "meituan-longcat/LongCat-Flash-Chat",
           "capabilities": [
             "tools"
-          ]
+          ],
+          "size_b": 561.9
         },
         {
           "name": "x-ai/grok-code-fast-1",
@@ -6584,7 +6629,8 @@
             "tools",
             "reasoning",
             "eu-endpoint"
-          ]
+          ],
+          "size_b": 228.7
         },
         {
           "name": "qwen/qwen3.5-122b-a10b",
@@ -6625,7 +6671,8 @@
             "tools",
             "reasoning",
             "eu-endpoint"
-          ]
+          ],
+          "size_b": 685.4
         },
         {
           "name": "qwen/qwen3-vl-235b-a22b-thinking",
@@ -6682,7 +6729,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 228.7
         },
         {
           "name": "nex-agi/deepseek-v3.1-nex-n1",
@@ -6693,7 +6741,8 @@
           "hf_id": "nex-agi/DeepSeek-V3.1-Nex-N1",
           "capabilities": [
             "tools"
-          ]
+          ],
+          "size_b": 671.1
         },
         {
           "name": "deepseek/deepseek-v3.2-exp",
@@ -6891,7 +6940,8 @@
             "tools",
             "reasoning",
             "eu-endpoint"
-          ]
+          ],
+          "size_b": 358.3
         },
         {
           "name": "z-ai/glm-4.6",
@@ -6903,7 +6953,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 356.8
         },
         {
           "name": "xiaomi/mimo-v2-omni",
@@ -6951,7 +7002,8 @@
           "hf_id": "moonshotai/Kimi-K2-Instruct-0905",
           "capabilities": [
             "tools"
-          ]
+          ],
+          "size_b": 1026.5
         },
         {
           "name": "mistralai/mistral-medium-3.1",
@@ -7070,7 +7122,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 684.5
         },
         {
           "name": "undi95/remm-slerp-l2-13b",
@@ -7105,7 +7158,8 @@
             "tools",
             "reasoning",
             "eu-endpoint"
-          ]
+          ],
+          "size_b": 1058.1
         },
         {
           "name": "google/gemini-3.1-flash-image-preview",
@@ -7261,7 +7315,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 358.3
         },
         {
           "name": "nvidia/llama-3.1-nemotron-ultra-253b-v1",
@@ -7355,7 +7410,8 @@
             "tools",
             "reasoning",
             "eu-endpoint"
-          ]
+          ],
+          "size_b": 684.5
         },
         {
           "name": "z-ai/glm-5",
@@ -7367,7 +7423,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 753.9
         },
         {
           "name": "openai/gpt-5.4-mini",
@@ -8733,7 +8790,8 @@
           "capabilities": [
             "tools",
             "reasoning"
-          ]
+          ],
+          "size_b": 199.4
         },
         {
           "name": "arcee-ai/trinity-large-preview:free",
@@ -9034,7 +9092,8 @@
           "capabilities": [
             "tools"
           ],
-          "hf_id": "moonshotai/Kimi-K2-Instruct-0905"
+          "hf_id": "moonshotai/Kimi-K2-Instruct-0905",
+          "size_b": 1026.5
         },
         {
           "name": "Llama 4 Scout (17Bx16E) 128k",

scripts/fetch-providers.js

@@ -52,7 +52,26 @@ function updateProviderModels(providers, providerName, models) {
     console.warn(`  ⚠  Provider "${providerName}" not found in providers.json – skipping.`);
     return false;
   }
-  provider.models = models;
+
+  // Smart merge: preserve existing metadata (size_b, hf_id, capabilities) if missing in new data
+  const existingMap = new Map((provider.models || []).map(m => [m.name, m]));
+  
+  provider.models = models.map(newModel => {
+    const existing = existingMap.get(newModel.name);
+    if (!existing) return newModel;
+
+    return {
+      ...existing, // Start with existing metadata
+      ...newModel, // Overwrite with new prices/type
+      // But preserve these if newModel doesn't have them
+      size_b: newModel.size_b || existing.size_b,
+      hf_id: newModel.hf_id || existing.hf_id,
+      capabilities: (newModel.capabilities && newModel.capabilities.length > 0) 
+        ? newModel.capabilities 
+        : existing.capabilities,
+    };
+  });
+
   return true;
 }
 
@@ -142,7 +161,7 @@ async function fetchHFSize(hfId) {
   const token = process.env.HF_TOKEN;
   const headers = token ? { Authorization: `Bearer ${token}` } : {};
   try {
-    const data = await getJson(`https://huggingface.co/api/models/${hfId}`, { headers });
+    const data = await getJson(`https://huggingface.co/api/models/${hfId}`, { headers, retries: 1 });
     // Check various common metadata locations for total parameters
     let params = data.safetensors?.total || data.config?.total_parameters || data.config?.model_type_params;
     if (!params && data.cardData?.model_details?.parameters) {
@@ -164,7 +183,13 @@ const EMBEDDER_KEYWORDS = ['embed', 'bge', 'gte', 'e5', 'stella', 'minilm', 'mul
 const MANUAL_HF_ID_MAP = {
   'all minilm l12 v2': 'sentence-transformers/all-MiniLM-L12-v2',
   'whisper v3': 'openai/whisper-large-v3',
-  'whisper-large-v3': 'openai/whisper-large-v3',
+  'whisper large v3': 'openai/whisper-large-v3',
+  'step 3 5 flash': 'stepfun-ai/Step-3.5-Flash',
+  'bge m3': 'BAAI/bge-m3',
+};
+
+const MANUAL_SIZE_MAP = {
+  'BAAI/bge-m3': 0.57,
 };
 
 // Propagate capabilities and size from benchmarks, OpenRouter, or HF Hub to all other providers' models.
@@ -210,13 +235,16 @@ async function propagateExtraData(data) {
 
       // 1. STRUCTURED LOOKUP: Match size by hf_id if available (Benchmark gold-standard)
       if (!model.size_b && model.hf_id) {
-        const size = hfIdToSize.get(model.hf_id.toLowerCase());
-        if (size) {
-          model.size_b = size;
+        if (MANUAL_SIZE_MAP[model.hf_id]) {
+          model.size_b = MANUAL_SIZE_MAP[model.hf_id];
           propagatedSize++;
+        } else {
+          const size = hfIdToSize.get(model.hf_id.toLowerCase());
+          if (size) { model.size_b = size; propagatedSize++; }
         }
       }
 
+
       // 2. AUTO-TAG image-gen and embedding models
       if (model.type === 'image' && (!model.capabilities || !model.capabilities.length)) {
         model.capabilities = ['image-gen'];
@@ -271,27 +299,40 @@ async function propagateExtraData(data) {
     }
   }
 
-  // 7. HUB API: Inspect technical metadata (Limit 30 unique IDs to avoid long startup)
-  const uniqueIds = [...new Set(hfLookupQueue.map(m => m.hf_id || m.name).filter(id => id.includes('/')))].slice(0, 30);
+  // 7. HUB API: Inspect technical metadata (Limit 200 unique IDs to ensure better coverage)
+  const uniqueIds = [...new Set(hfLookupQueue.map(m => m.hf_id || m.name).filter(id => id.includes('/')))].slice(0, 200);
   if (uniqueIds.length > 0) {
-    process.stdout.write(`  HF Hub: technical metadata inspection for ${uniqueIds.length} models... `);
+    console.log(`\n  HF Hub: technical metadata inspection for ${uniqueIds.length} models...`);
     const idToSize = new Map();
-    await Promise.all(uniqueIds.map(async (id) => {
+    
+    // Process sequentially with small delay to avoid 429 rate limits
+    for (let i = 0; i < uniqueIds.length; i++) {
+      const id = uniqueIds[i];
+      process.stdout.write(`    [${i + 1}/${uniqueIds.length}] ${id.padEnd(50)} `);
       const size = await fetchHFSize(id);
-      if (size) idToSize.set(id, size);
-    }));
+      if (size) {
+        idToSize.set(id, size);
+        process.stdout.write(`✓ ${size}B\n`);
+      } else {
+        process.stdout.write(`✗\n`);
+      }
+      await new Promise(r => setTimeout(r, 50)); // Tiny delay
+    }
+
     for (const model of hfLookupQueue) {
       if (!model.size_b) {
-        const size = idToSize.get(model.hf_id || model.name);
+        const id = model.hf_id || model.name;
+        const size = idToSize.get(id);
         if (size) {
           model.size_b = size;
           hfSizeFetched++;
         }
       }
     }
-    console.log(`✓ ${hfSizeFetched} sizes found`);
+    console.log(`  ✓ Total ${hfSizeFetched} new sizes from HF metadata`);
   }
 
+
   if (autoTagged > 0) console.log(`Auto-tagged ${autoTagged} image-gen/embedding models.`);
   if (propagatedCaps > 0) console.log(`Propagated capabilities to ${propagatedCaps} models.`);
   if (propagatedSize + hfSizeFetched > 0) console.log(`Enriched size data for ${propagatedSize + hfSizeFetched} models.`);