✨ Add GPT 5.5 + Codex

results/swe-bench-pro--ansible-gpt-5-5-codex.json

@@ -0,0 +1,60 @@
+{
+    "benchmark": {
+        "name": "swe-bench-pro--ansible",
+        "repo": "ScaleAI/SWE-bench_Pro",
+        "num_tasks": 96,
+        "url": "https://huggingface.co/datasets/ScaleAI/SWE-bench_Pro"
+    },
+    "harness": {
+        "name": "Codex",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/openai/codex"
+    },
+    "model": {
+        "name": "GPT 5.5",
+        "repo": "GPT 5.5",
+        "is_oss": true,
+        "num_params": 9700,
+        "precision": "bf16",
+        "url": "https://openai.com/index/introducing-gpt-5-5"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "path": null,
+            "name": "scale-ai/swe-bench-pro",
+            "version": null,
+            "ref": "sha256:88411d32ff27e53a4c1a7e29f0c2aeba180c8e5d60f221cab5ed56325f33549d",
+            "registry_url": null,
+            "registry_path": null,
+            "overwrite": false,
+            "download_dir": null,
+            "task_names": [
+                "*ansible*"
+            ],
+            "exclude_task_names": null,
+            "n_tasks": null
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 96,
+        "n_errors": 0,
+        "score": 0.604,
+        "n_input_tokens": 198924339,
+        "n_cache_tokens": 189578624,
+        "n_output_tokens": 1560836,
+        "n_total_tokens": 390063799,
+        "agent_time_seconds": 32914,
+        "total_time_seconds": 39468,
+        "cost_usd": 188.34296700000002,
+        "mean_input_tokens_per_task": 2072128,
+        "mean_cache_tokens_per_task": 1974777,
+        "mean_output_tokens_per_task": 16258,
+        "mean_tokens_per_task": 4063164,
+        "mean_cost_usd_per_task": 1.96,
+        "mean_total_time_seconds_per_task": 411,
+        "mean_agent_time_seconds_per_task": 342
+    }
+}

results/swe-bench-verified-gpt-5-5-codex.json

@@ -0,0 +1,58 @@
+{
+    "benchmark": {
+        "name": "swe-bench-verified",
+        "repo": "SWE-bench/SWE-bench_Verified",
+        "num_tasks": 500,
+        "url": "https://huggingface.co/datasets/SWE-bench/SWE-bench_Verified"
+    },
+    "harness": {
+        "name": "Codex",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/openai/codex"
+    },
+    "model": {
+        "name": "GPT 5.5",
+        "repo": "GPT 5.5",
+        "is_oss": true,
+        "num_params": 9700,
+        "precision": "bf16",
+        "url": "https://openai.com/index/introducing-gpt-5-5"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "path": null,
+            "name": "swe-bench/swe-bench-verified",
+            "version": null,
+            "ref": "sha256:235d6032d549851a936db3b5fe08807c4d385c12ee10e7be9c9786a1ff60563c",
+            "registry_url": null,
+            "registry_path": null,
+            "overwrite": false,
+            "download_dir": null,
+            "task_names": null,
+            "exclude_task_names": null,
+            "n_tasks": null
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 500,
+        "n_errors": 0,
+        "score": 0.798,
+        "n_input_tokens": 415795756,
+        "n_cache_tokens": 386352384,
+        "n_output_tokens": 3431224,
+        "n_total_tokens": 805579364,
+        "agent_time_seconds": 92588,
+        "total_time_seconds": 141637,
+        "cost_usd": 443.32977200000016,
+        "mean_input_tokens_per_task": 831591,
+        "mean_cache_tokens_per_task": 772704,
+        "mean_output_tokens_per_task": 6862,
+        "mean_tokens_per_task": 1611158,
+        "mean_cost_usd_per_task": 0.89,
+        "mean_total_time_seconds_per_task": 283,
+        "mean_agent_time_seconds_per_task": 185
+    }
+}