Leading or double-barrelled questions:
23 results above zero · % of tasks, lower is better. Choose a model to highlight it.Clear highlight
- 5 Qwen4-MaxAlibaba 11.1%
- 5 DeepSeek-V4.1-Flash-1017DeepSeek 11.1%
- 5 GPT-6.1 TerraOpenAI 11.1%
- 5 GPT-6 LunaOpenAI 11.1%
- 5 Particle-1Particle AI 11.1%
- 10 Claude Fable 5.2Anthropic 22.2%
- 10 Kimi K3.1 ProMoonshot AI 22.2%
- 12 Claude Haiku 5.5Anthropic 33.3%
- 12 Muse EmberMeta 33.3%
- 12 Mistral Large 4Mistral AI 33.3%
- 15 Gemini 3.1 Pro PreviewGoogle 44.4%
- 15 Gemini 4 HydrogenGoogle 44.4%
- 15 Muse WonderMeta 44.4%
- 15 MiniMax M3.1 FlashMiniMax 44.4%
- 15 GLM-5.3 FlashX LegalZ.ai 44.4%
4 models had none: Claude Opus 5.5, GPT-6.1 Sol, GPT-6.2 Astra, GPT-6.2 Sol.