add missing model results and fix kimi_k26 multimodal

.eval_results/deepseek_v4_flash.yaml

@@ -0,0 +1,18 @@
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: general
+  value: 57.8
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 161 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: multi_turn
+  value: 57.9
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 38 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard

.eval_results/deepseek_v4_pro.yaml

@@ -0,0 +1,18 @@
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: general
+  value: 58.4
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 161 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: multi_turn
+  value: 65.8
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 38 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard

.eval_results/glm51.yaml

@@ -0,0 +1,18 @@
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: general
+  value: 62.7
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 161 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: multi_turn
+  value: 60.5
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 38 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard

.eval_results/kimi_k26.yaml

@@ -7,6 +7,15 @@
   source:
     url: https://claw-eval.github.io
     name: Claw-Eval Leaderboard
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: multimodal
+  value: 18.8
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 101 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard
 - dataset:
     id: claw-eval/Claw-Eval
     task_id: multi_turn

.eval_results/qwen3_6_plus.yaml

@@ -0,0 +1,18 @@
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: general
+  value: 57.1
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 161 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard
+- dataset:
+    id: claw-eval/Claw-Eval
+    task_id: multi_turn
+  value: 65.8
+  date: '2026-04-23'
+  notes: Pass³% | N=3 | 38 tasks
+  source:
+    url: https://claw-eval.github.io
+    name: Claw-Eval Leaderboard