Writing
Short writing briefs. No code can score them, so two judge models rate each piece from 0 to 10 against a rubric.
This category is shown on its own and does not count toward Overall. How the battery works.
Ranking
No model has a programmatic score in this category. Models are ranked by the average of the two judges' mean scores, from 0 to 10. Only models that both judges scored are ranked here. These judge scores are never part of Overall.
| Rank | Model | Judges (0 to 10) | Items | Judge A mean | Judge B mean | Agreement |
|---|---|---|---|---|---|---|
| 1 | gemma-4-31B-it | 7.0 | 8 | 7.4openai/gpt-5.6-luna | 6.6anthropic/claude-haiku-5.5 | medium |
| 1 | Muse Glimmer 30B | 7.0 | 8 | 7.4openai/gpt-5.6-luna | 6.6anthropic/claude-haiku-5.5 | high |
| 3 | gemma-3-27b-it | 6.9 | 8 | 7.5openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | medium |
| 4 | gemma-4-26B-A4B-it | 6.8 | 8 | 6.9openai/gpt-5.6-luna | 6.8anthropic/claude-haiku-5.5 | high |
| 4 | Qwen3.5-27B | 6.8 | 8 | 7.4openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | medium |
| 4 | Qwen3.5-35B-A3B | 6.8 | 8 | 7.0openai/gpt-5.6-luna | 6.6anthropic/claude-haiku-5.5 | high |
| 7 | Gemma 4 12B IT | 6.8 | 8 | 7.4openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | high |
| 7 | Gemma4 E2B IT | 6.8 | 8 | 7.3openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | medium |
| 9 | gpt-oss-120b | 6.7 | 8 | 6.3anthropic/claude-haiku-5.5 | 7.1deepseek/deepseek-v4.1-flash | high |
| 10 | Sage-A1.5-Q4_K_M | 6.6 | 8 | 7.1openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | medium |
| 11 | Mistral Small 3.2 24B Instruct (2506) | 6.5 | 8 | 6.8openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | medium |
| 12 | Granite4.2 30B | 6.4 | 8 | 6.6openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | high |
| 13 | Flint 40B A2B | 6.4 | 8 | 6.5openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | high |
| 13 | Olmo 3.1 32B Instruct | 6.4 | 8 | 6.8openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 13 | Qwen3.5-9B | 6.4 | 8 | 6.6openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | medium |
| 16 | Gemma3n E2B IT | 6.3 | 8 | 6.8openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | medium |
| 16 | Granite4.2 8B | 6.3 | 8 | 6.4openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | high |
| 16 | Ministral 3 14B Instruct (2512) | 6.3 | 8 | 6.6openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 16 | Qwen3.6-27B | 6.3 | 8 | 6.5openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | high |
| 16 | Qwen3.6-35B-A3B | 6.3 | 8 | 6.6openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 21 | Llama-3.3-70B-Instruct | 6.3 | 8 | 6.6openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 21 | Medgemma 27B ITfine-tune of Gemma 3medical model | 6.3 | 8 | 6.3openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | high |
| 21 | Qwen2.5-32B-Instruct | 6.3 | 8 | 6.4openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | medium |
| 21 | Qwen3-14B | 6.3 | 8 | 6.5openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 21 | Qwen3-30B-A3B-Instruct-2507 | 6.3 | 8 | 6.8openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 21 | Qwen3-Coder-Nextcode model | 6.3 | 8 | 6.8openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 27 | Gemma 3n E4B IT | 6.2 | 8 | 6.8openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | medium |
| 27 | gemma-3-12b-it | 6.2 | 8 | 6.5openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 27 | GLM-4.5-Air | 6.2 | 8 | 6.4openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 27 | Nemotron 3.5 Bolt E4B | 6.2 | 8 | 6.4openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 27 | Nemotron3 33B | 6.2 | 8 | 6.4openai/gpt-5.6-luna | 6.0anthropic/claude-haiku-5.5 | high |
| 27 | Ornith 9B | 6.2 | 8 | 6.5openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 27 | Qwen3-32B | 6.2 | 8 | 6.1openai/gpt-5.6-luna | 6.3anthropic/claude-haiku-5.5 | high |
| 27 | Qwen3-4B-Instruct-2507 | 6.2 | 8 | 6.6openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | medium |
| 35 | Devstral Small 2 24B Instruct (2512)code model | 6.1 | 8 | 6.5openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 35 | Ornith-1.5-9B | 6.1 | 8 | 6.4openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 35 | Qwen3.8-27B | 6.1 | 8 | 6.1openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | high |
| 38 | Meta-Llama-3-8B-Instruct | 6.1 | 8 | 6.5openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 38 | Ministral 3 8B Instruct 2512 | 6.1 | 8 | 6.3openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 40 | QwQ 32B | 6.0 | 8 | 6.3openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 41 | Gemma4 E4B IT | 5.9 | 8 | 5.8openai/gpt-5.6-luna | 6.1anthropic/claude-haiku-5.5 | high |
| 42 | Devstral 24B Small 2505code model | 5.9 | 8 | 5.9openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 42 | Laguna XS 2.1 | 5.9 | 8 | 6.1openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 42 | Translategemma 12B ITfine-tune of Gemma 3translation model | 5.9 | 8 | 6.0openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 45 | DeepSeek-R1-Distill-Qwen-32B | 5.8 | 8 | 5.8openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 45 | GLM-4.7-Flash | 5.8 | 8 | 6.0openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 45 | Granite3.1 Dense 8B Instruct | 5.8 | 8 | 6.1openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 45 | Qwen3.5-4B | 5.8 | 8 | 5.9openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | medium |
| 49 | Cogito 14B V1 Preview Qwenfine-tune of Qwen 2.5 | 5.8 | 8 | 5.8openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 49 | NVIDIA Nemotron 3 Nano 30B-A3B | 5.8 | 8 | 5.8openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | medium |
| 49 | Qwen3-VL-30B-A3B-Instruct | 5.8 | 8 | 5.6openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | high |
| 52 | Nanbeige_Nanbeige4.3-33B | 5.7 | 8 | 6.1openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 53 | Command R7B Arabic 7B 02 2025 | 5.7 | 8 | 5.9openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 53 | gemma-2-9b-it | 5.7 | 8 | 5.9openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 53 | Granite4 Small H | 5.7 | 8 | 5.9openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 53 | Granite4.1 30B | 5.7 | 8 | 5.8openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 53 | Lfm2 24B | 5.7 | 8 | 6.0openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 53 | Llama 4 Scout 17B-16E Instruct | 5.7 | 8 | 5.5openai/gpt-5.6-luna | 5.9anthropic/claude-haiku-5.5 | medium |
| 53 | Ministral 3 3B Instruct 2512 | 5.7 | 8 | 6.1openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 53 | phi-4 | 5.7 | 8 | 5.6openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 53 | Qwen3-4B-Thinking-2507 | 5.7 | 8 | 6.0openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 62 | Nemotron Cascade 2 30B A3B | 5.6 | 8 | 5.9openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 62 | Qwen3-4B | 5.6 | 8 | 5.6openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 64 | Aya Expanse 8B | 5.6 | 8 | 5.8openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 64 | gemma-3-4b-it | 5.6 | 8 | 5.6openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 64 | Mistral Small3.1 24B Instruct 2503 | 5.6 | 8 | 5.6openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 64 | Phi-3-mini-4k-instruct | 5.6 | 8 | 5.5openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 64 | Qwen2.5-Coder-32B-Instructcode model | 5.6 | 8 | 5.4openai/gpt-5.6-luna | 5.8anthropic/claude-haiku-5.5 | high |
| 64 | Tulu3 8Bfine-tune of Llama 3.1 | 5.6 | 8 | 5.8openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 70 | Cogito 8B V1 Preview Llamafine-tune of Llama 3.1 | 5.5 | 8 | 5.6openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 70 | Granite4.1 8B | 5.5 | 8 | 5.6openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 70 | Llama-3.1-8B-Instruct | 5.5 | 8 | 5.8openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 70 | Llama-3.2-3B-Instruct | 5.5 | 8 | 5.8openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 70 | Mistral Small 22B Instruct 2409 | 5.5 | 8 | 5.9openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | medium |
| 70 | Olmo2 13B 1124 Instruct | 5.5 | 8 | 5.8openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 70 | Qwen3-8B | 5.5 | 8 | 5.8openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 70 | Qwen3-VL-32B-Instruct | 5.5 | 8 | 5.5openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | medium |
| 78 | Command R 35B 08 2024 | 5.4 | 8 | 5.6openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 78 | Gemma2 27B Instruct | 5.4 | 8 | 5.4openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 78 | Hermes-3-Llama-3.1-8B | 5.4 | 8 | 5.8openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 78 | Mistral Small 24B Instruct 2501 | 5.4 | 8 | 5.6openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 78 | Nemotron 3.5 Lightning 30B A3B | 5.4 | 8 | 5.6openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 78 | Qwen2.5-14B-Instruct | 5.4 | 8 | 5.5openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 78 | Qwen2.5-Coder-14B-Instructcode model | 5.4 | 8 | 5.5openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 78 | Qwen3-VL-8B-Instruct | 5.4 | 8 | 5.8openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 78 | Solar 10.7B Instruct V1 | 5.4 | 8 | 5.8openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 87 | DeepSeek-R1-0528-Qwen3-8B | 5.4 | 8 | 5.5openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 87 | Exaone3.5 32B Instruct | 5.4 | 8 | 5.3openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 87 | Qwen3-30B-A3B | 5.4 | 8 | 5.4openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 90 | Command R7B (12-2024) | 5.3 | 8 | 5.6openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 90 | Granite3.2 8B Instruct | 5.3 | 8 | 5.4openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 92 | Deepcoder 14B Previewfine-tune of DeepSeek-R1-Distill-Qwen 14Bcode model | 5.3 | 8 | 5.0openai/gpt-5.6-luna | 5.5anthropic/claude-haiku-5.5 | high |
| 92 | Falcon3 10B Instruct | 5.3 | 8 | 5.3openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 92 | Granite3.3 8B | 5.3 | 8 | 5.4openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 92 | Mistral 7B Instruct V0.3 | 5.3 | 8 | 5.4openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 92 | Qwen2.5-VL-32B-Instruct | 5.3 | 8 | 4.9openai/gpt-5.6-luna | 5.6anthropic/claude-haiku-5.5 | high |
| 92 | SmolLM3-3B | 5.3 | 8 | 5.5openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 98 | Granite3.3 2B | 5.2 | 8 | 5.0openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 98 | Granite4 3B | 5.2 | 8 | 5.4openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 98 | Lfm3 VL 2B | 5.2 | 8 | 5.5openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 98 | Llama2 13B Chat | 5.2 | 8 | 5.6openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 98 | medgemma-4b-itmedical model | 5.2 | 8 | 5.4openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 98 | Olmo-3-7B-Instruct | 5.2 | 8 | 5.0openai/gpt-5.6-luna | 5.4anthropic/claude-haiku-5.5 | high |
| 98 | Phi3 14B Medium 128K Instruct | 5.2 | 8 | 5.5openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | medium |
| 105 | Gemma2 2B Instruct | 5.1 | 8 | 5.1openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | medium |
| 105 | Gemma3 1B IT | 5.1 | 8 | 5.4openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 105 | granite-4.1-3b | 5.1 | 8 | 5.3openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 105 | Granite4.2 3B | 5.1 | 8 | 5.0openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 105 | Qwen3-Coder-30B-A3B-Instructcode model | 5.1 | 8 | 5.0openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 105 | Qwen3-VL-4B-Instruct | 5.1 | 8 | 5.5openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | medium |
| 105 | Spark-X2.6-4B | 5.1 | 8 | 5.1openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 112 | Aya Expanse 32B | 5.1 | 8 | 5.3openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 112 | Cogito 3B V1 Preview Llamafine-tune of Llama 3.2 | 5.1 | 8 | 4.9openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 112 | Deepseek V2 16B Lite Chat | 5.1 | 8 | 5.3openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 112 | Exaone3.5 2.4B Instruct | 5.1 | 8 | 4.9openai/gpt-5.6-luna | 5.3anthropic/claude-haiku-5.5 | high |
| 112 | Falcon3 3B Instruct | 5.1 | 8 | 5.3openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 112 | Granite4 Tiny H | 5.1 | 8 | 5.0openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 112 | Medgemma1.5 4B ITfine-tune of Gemma 3medical model | 5.1 | 8 | 5.4openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 119 | Codegemma 7B Instruct V1.1code model | 5.0 | 8 | 5.0openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 119 | DeepSeek-R1-Distill-Qwen-14B | 5.0 | 8 | 5.1openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 119 | dolphin-2.9.1-yi-1.5-34b | 5.0 | 8 | 4.9openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 119 | Falcon3 7B Instruct | 5.0 | 8 | 5.3openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 123 | Granite4 Micro H | 4.9 | 8 | 5.0openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 124 | EXAONE-3.5-7.8B-Instruct | 4.9 | 8 | 4.6openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 124 | Gemma 7B Instruct V1.1 | 4.9 | 8 | 5.1openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 124 | Mixtral 8x7B Instruct V0.1 | 4.9 | 8 | 4.6openai/gpt-5.6-luna | 5.1anthropic/claude-haiku-5.5 | high |
| 124 | Olmo-3-7B-Think | 4.9 | 8 | 4.8openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 124 | Olmo2 7B 1124 Instruct | 4.9 | 8 | 4.9openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 124 | YI 34B Chat V1.5 | 4.9 | 8 | 4.9openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 124 | YI 6B Chat V1.5 | 4.9 | 8 | 4.9openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | medium |
| 131 | Deepseek R1 8B Llama Distill | 4.8 | 8 | 4.9openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 131 | Granite3.2 2B Instruct | 4.8 | 8 | 4.8openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 131 | Lfm3 1.7B | 4.8 | 8 | 4.6openai/gpt-5.6-luna | 5.0anthropic/claude-haiku-5.5 | high |
| 131 | Mistral-7B-Instruct-v0.2 | 4.8 | 8 | 4.8openai/gpt-5.6-luna | 4.9anthropic/claude-haiku-5.5 | high |
| 135 | Codestral 22B V0.1code model | 4.8 | 8 | 4.9openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 135 | DeepSeek-R1-Distill-Qwen-7B | 4.8 | 8 | 5.0openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 135 | Internlm2 7B Chat V2.5 | 4.8 | 8 | 4.9openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 138 | Aya 8B 23 | 4.7 | 8 | 4.6openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 138 | Glm4 9B Chat | 4.7 | 8 | 4.9openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 138 | Granite3 Dense 8B Instruct | 4.7 | 8 | 4.8openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 138 | Qwen2.5-7B-Instruct | 4.7 | 8 | 4.6openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 138 | Qwen2.5-Coder-7B-Instructcode model | 4.7 | 8 | 4.9openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 138 | Qwen3 1.7B | 4.7 | 8 | 4.9openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 144 | gpt-oss-20b | 4.6 | 8 | 4.1anthropic/claude-haiku-5.5 | 5.1deepseek/deepseek-v4.1-flash | high |
| 144 | Granite3.1 Dense 2B Instruct | 4.6 | 8 | 4.5openai/gpt-5.6-luna | 4.8anthropic/claude-haiku-5.5 | high |
| 144 | Phi-3.5-mini-instruct | 4.6 | 8 | 4.8openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 144 | Qwen2-VL-7B-Instruct | 4.6 | 8 | 4.8openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 144 | Qwen2.5-3B-Instruct | 4.6 | 8 | 4.6openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 144 | Qwen3.5 2B | 4.6 | 8 | 4.8openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 150 | Apertus-8B-Instruct-2509 | 4.6 | 8 | 4.9openai/gpt-5.6-luna | 4.3anthropic/claude-haiku-5.5 | high |
| 150 | Llama3.2 1B Instruct | 4.6 | 8 | 5.1openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 150 | Phi-4-mini-instruct | 4.6 | 8 | 4.5openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 150 | Qwen 14B Chat V1.5 | 4.6 | 8 | 4.8openai/gpt-5.6-luna | 4.4anthropic/claude-haiku-5.5 | high |
| 150 | Sailor2 8B Chatfine-tune of Qwen 2.5 | 4.6 | 8 | 5.0openai/gpt-5.6-luna | 4.1anthropic/claude-haiku-5.5 | high |
| 155 | Codellama 7B Instructcode model | 4.5 | 8 | 4.6openai/gpt-5.6-luna | 4.4anthropic/claude-haiku-5.5 | high |
| 155 | Phi3 3.8B Mini 128K Instruct | 4.5 | 8 | 4.5openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 155 | Qwen2.5-VL-7B-Instruct | 4.5 | 8 | 4.9openai/gpt-5.6-luna | 4.1anthropic/claude-haiku-5.5 | high |
| 158 | DeepSeek-Coder-V2-Lite-Instructcode model | 4.4 | 8 | 4.4openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 158 | Granite3.1 Moe 3B Instruct | 4.4 | 8 | 4.3openai/gpt-5.6-luna | 4.6anthropic/claude-haiku-5.5 | high |
| 158 | Qwen2.5-VL-3B-Instruct | 4.4 | 8 | 4.9openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 158 | YI 9B Chat V1.5 | 4.4 | 8 | 4.4openai/gpt-5.6-luna | 4.5anthropic/claude-haiku-5.5 | high |
| 162 | Granite3 Dense 2B Instruct | 4.4 | 8 | 4.4openai/gpt-5.6-luna | 4.4anthropic/claude-haiku-5.5 | high |
| 162 | Mistral Nemo 12B Instruct 2407 | 4.4 | 8 | 4.4openai/gpt-5.6-luna | 4.4anthropic/claude-haiku-5.5 | high |
| 164 | Qwen2 Math 7B Instructmath model | 4.3 | 8 | 4.6openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 164 | Qwen2.5 1.5B Instruct | 4.3 | 8 | 4.6openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 164 | Qwen3 VL 2B Instruct | 4.3 | 8 | 5.0openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | medium |
| 167 | Granite4 1B H | 4.3 | 8 | 4.9openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | high |
| 167 | Hermes3 3B Llama3.2fine-tune of Llama 3.2 | 4.3 | 8 | 4.5openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 167 | Lfm2.5 8B A1B | 4.3 | 8 | 4.6openai/gpt-5.6-luna | 3.9anthropic/claude-haiku-5.5 | high |
| 167 | Llama-2-7b-chat-hf | 4.3 | 8 | 4.5openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 167 | Qwen2.5-Coder-3B-Instructcode model | 4.3 | 8 | 4.8openai/gpt-5.6-luna | 3.8anthropic/claude-haiku-5.5 | high |
| 172 | Codellama 13B Instructcode model | 4.2 | 8 | 4.6openai/gpt-5.6-luna | 3.8anthropic/claude-haiku-5.5 | high |
| 172 | Lfm2.5 Thinking 1.2B | 4.2 | 8 | 4.6openai/gpt-5.6-luna | 3.8anthropic/claude-haiku-5.5 | high |
| 172 | Smollm2 1.7B Instruct | 4.2 | 8 | 4.8openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | high |
| 175 | Exaone Deep 7.8B | 4.1 | 8 | 5.0openai/gpt-5.6-luna | 3.3anthropic/claude-haiku-5.5 | medium |
| 175 | Nemotron Mini 4B Instruct | 4.1 | 8 | 4.4openai/gpt-5.6-luna | 3.9anthropic/claude-haiku-5.5 | high |
| 175 | Stablelm Zephyr 3Bfine-tune of StableLM 3B | 4.1 | 8 | 4.4openai/gpt-5.6-luna | 3.9anthropic/claude-haiku-5.5 | high |
| 175 | Stablelm2 12B Chat | 4.1 | 8 | 4.4openai/gpt-5.6-luna | 3.9anthropic/claude-haiku-5.5 | high |
| 179 | Codellama 34B Instructcode model | 4.1 | 8 | 4.5openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | high |
| 179 | Mathstral 7B V0.1fine-tune of Mistral 7Bmath model | 4.1 | 8 | 4.8openai/gpt-5.6-luna | 3.4anthropic/claude-haiku-5.5 | medium |
| 179 | Qwen 7B Chat V1.5 | 4.1 | 8 | 4.3openai/gpt-5.6-luna | 3.9anthropic/claude-haiku-5.5 | high |
| 182 | Deepseek Llm 7B Chat | 4.0 | 8 | 4.4openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | high |
| 182 | Gemma 2B Instruct V1.1 | 4.0 | 8 | 4.0openai/gpt-5.6-luna | 4.0anthropic/claude-haiku-5.5 | high |
| 182 | Granite Code 8B Instructcode model | 4.0 | 8 | 4.6openai/gpt-5.6-luna | 3.4anthropic/claude-haiku-5.5 | high |
| 185 | Marco O1 7Bfine-tune of Qwen 2 7B | 3.9 | 8 | 4.5openai/gpt-5.6-luna | 3.4anthropic/claude-haiku-5.5 | medium |
| 186 | Codegeex4 9B Allcode model | 3.9 | 8 | 4.4openai/gpt-5.6-luna | 3.4anthropic/claude-haiku-5.5 | high |
| 186 | Granite3 Moe 3B Instruct | 3.9 | 8 | 4.3openai/gpt-5.6-luna | 3.5anthropic/claude-haiku-5.5 | high |
| 186 | Granite3.1 Moe 1B Instruct | 3.9 | 8 | 4.3openai/gpt-5.6-luna | 3.5anthropic/claude-haiku-5.5 | high |
| 186 | Magistral Small (2506) | 3.9 | 8 | 4.1openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | high |
| 186 | Qwen2 1.5B Instruct | 3.9 | 8 | 4.1openai/gpt-5.6-luna | 3.6anthropic/claude-haiku-5.5 | high |
| 186 | Qwen2.5 Coder 1.5B Instructcode model | 3.9 | 8 | 4.5openai/gpt-5.6-luna | 3.3anthropic/claude-haiku-5.5 | medium |
| 192 | YI Coder 9B Chatcode model | 3.8 | 8 | 3.8openai/gpt-5.6-luna | 3.9anthropic/claude-haiku-5.5 | high |
| 193 | Smollm2 360M Instruct | 3.7 | 8 | 4.5openai/gpt-5.6-luna | 2.9anthropic/claude-haiku-5.5 | medium |
| 194 | Dolphin Phi 2.7B V2.6fine-tune of Phi-2 | 3.6 | 8 | 4.3openai/gpt-5.6-luna | 2.9anthropic/claude-haiku-5.5 | medium |
| 194 | Smallthinker 3B Previewfine-tune of Qwen 2.5 3B | 3.6 | 8 | 4.3openai/gpt-5.6-luna | 2.9anthropic/claude-haiku-5.5 | high |
| 196 | Internlm2 1.8B Chat V2.5 | 3.5 | 8 | 4.1openai/gpt-5.6-luna | 2.9anthropic/claude-haiku-5.5 | medium |
| 196 | Openthinker 7B V2fine-tune of Qwen 2.5 | 3.5 | 8 | 4.1openai/gpt-5.6-luna | 2.9anthropic/claude-haiku-5.5 | high |
| 198 | Deepseek R1 1.5B Qwen Distill | 3.4 | 8 | 3.9openai/gpt-5.6-luna | 3.0anthropic/claude-haiku-5.5 | high |
| 198 | Granite3 Moe 1B Instruct | 3.4 | 8 | 3.6openai/gpt-5.6-luna | 3.3anthropic/claude-haiku-5.5 | high |
| 200 | Codeqwen 7B Chat V1.5code model | 3.4 | 8 | 4.0openai/gpt-5.6-luna | 2.8anthropic/claude-haiku-5.5 | high |
| 201 | Deepcoder 1.5B Previewfine-tune of DeepSeek-R1-Distill-Qwen 1.5Bcode model | 3.3 | 8 | 3.9openai/gpt-5.6-luna | 2.8anthropic/claude-haiku-5.5 | high |
| 201 | Qwen3 0.6B | 3.3 | 8 | 3.8openai/gpt-5.6-luna | 2.9anthropic/claude-haiku-5.5 | high |
| 203 | Falcon3 1B Instruct | 3.3 | 8 | 3.8openai/gpt-5.6-luna | 2.8anthropic/claude-haiku-5.5 | high |
| 203 | Qwen2 0.5B Instruct | 3.3 | 8 | 3.9openai/gpt-5.6-luna | 2.6anthropic/claude-haiku-5.5 | high |
| 203 | Qwen3 VL 2B Thinking | 3.3 | 8 | 3.4openai/gpt-5.6-luna | 3.1anthropic/claude-haiku-5.5 | high |
| 206 | Falcon2 11Bnot in the Overall ranking | 3.2 | 8 | 4.0openai/gpt-5.6-luna | 2.4anthropic/claude-haiku-5.5 | medium |
| 207 | Granite Code 3B Instructcode model | 3.1 | 8 | 3.9openai/gpt-5.6-luna | 2.4anthropic/claude-haiku-5.5 | medium |
| 207 | Phi 2.7B Chat V2 | 3.1 | 8 | 3.6openai/gpt-5.6-luna | 2.6anthropic/claude-haiku-5.5 | high |
| 207 | Sailor2 1B Chatfine-tune of Qwen 2.5 | 3.1 | 8 | 3.6openai/gpt-5.6-luna | 2.6anthropic/claude-haiku-5.5 | high |
| 207 | Stablelm2 1.6B Chat | 3.1 | 8 | 3.6openai/gpt-5.6-luna | 2.6anthropic/claude-haiku-5.5 | high |
| 211 | Phi4 Mini Reasoning 3.8B | 3.1 | 8 | 3.6openai/gpt-5.6-luna | 2.5anthropic/claude-haiku-5.5 | high |
| 212 | Granite4 350M H | 3.0 | 8 | 3.6openai/gpt-5.6-luna | 2.4anthropic/claude-haiku-5.5 | medium |
| 212 | Qwen2.5 0.5B Instruct | 3.0 | 8 | 3.5openai/gpt-5.6-luna | 2.5anthropic/claude-haiku-5.5 | high |
| 214 | Deepscaler 1.5B Previewfine-tune of DeepSeek-R1-Distill-Qwen 1.5Bmath model | 2.8 | 8 | 3.3openai/gpt-5.6-luna | 2.4anthropic/claude-haiku-5.5 | high |
| 214 | Exaone Deep 2.4B | 2.8 | 8 | 3.4openai/gpt-5.6-luna | 2.3anthropic/claude-haiku-5.5 | high |
| 214 | Qwen 4B Chat V1.5 | 2.8 | 8 | 3.6openai/gpt-5.6-luna | 2.0anthropic/claude-haiku-5.5 | medium |
| 217 | Qwen 1.8B Chat V1.5 | 2.7 | 8 | 3.3openai/gpt-5.6-luna | 2.1anthropic/claude-haiku-5.5 | high |
| 218 | Qwen2.5 Coder 0.5B Instructcode model | 2.5 | 8 | 3.1openai/gpt-5.6-luna | 1.9anthropic/claude-haiku-5.5 | high |
| 219 | YI Coder 1.5B Chatcode model | 2.4 | 8 | 3.3openai/gpt-5.6-luna | 1.6anthropic/claude-haiku-5.5 | medium |
| 220 | Opencoder 1.5B Instructcode model | 2.3 | 8 | 3.3openai/gpt-5.6-luna | 1.3anthropic/claude-haiku-5.5 | medium |
| 220 | Qwen 0.5B Chat V1.5 | 2.3 | 8 | 3.1openai/gpt-5.6-luna | 1.4anthropic/claude-haiku-5.5 | medium |
| 220 | Smollm2 135M Instruct | 2.3 | 8 | 3.0openai/gpt-5.6-luna | 1.5anthropic/claude-haiku-5.5 | medium |
| 223 | Qwen2 Math 1.5B Instructmath model | 2.1 | 8 | 3.0openai/gpt-5.6-luna | 1.1anthropic/claude-haiku-5.5 | medium |
| 224 | Stable Code 3B Instructcode model | 1.9 | 8 | 2.6openai/gpt-5.6-luna | 1.1anthropic/claude-haiku-5.5 | medium |
| 225 | Deepseek Coder 33B Instructcode model | 1.3 | 8 | 1.1openai/gpt-5.6-luna | 1.4anthropic/claude-haiku-5.5 | high |
| 226 | Deepseek Coder 6.7B Instructcode model | 1.1 | 8 | 1.3openai/gpt-5.6-luna | 1.0anthropic/claude-haiku-5.5 | high |
| 227 | Deepseek Coder 1.3B Instructcode model | 0.8 | 8 | 0.9openai/gpt-5.6-luna | 0.6anthropic/claude-haiku-5.5 | high |
| 228 | Codegemma 2B Code V1.1code model | 0.6 | 8 | 0.9openai/gpt-5.6-luna | 0.3anthropic/claude-haiku-5.5 | high |
A dash means not scored, which is different from 0. Agreement compares the two judges: high means a mean gap under 1.5 points, medium under 3, low otherwise; n/a means fewer than 3 items were scored by both.
Not ranked in this category
- Apertus-8B-Instruct-2509: withheld from the ranking (see its page)
- Falcon3 10B Instruct: withheld from the ranking (see its page)
- Falcon3 1B Instruct: withheld from the ranking (see its page)
- Falcon3 3B Instruct: withheld from the ranking (see its page)
- Falcon3 7B Instruct: withheld from the ranking (see its page)
- Internlm2 1.8B Chat V2.5: withheld from the ranking (see its page)
- Phi-4-reasoning 14B: withheld from the ranking (see its page)
- Phi-4-reasoning-plus 14B: withheld from the ranking (see its page)
- Qwen2 Math 1.5B Instruct: withheld from the ranking (see its page)
- Qwen2 Math 7B Instruct: withheld from the ranking (see its page)
The same ranking as JSON: writing.json.