Unsupported findings:
23 results above zero · % of tasks, lower is better. Choose a model to highlight it.Clear highlight
- 4 Claude Fable 5.2Anthropic 11.1%
- 4 Muse WonderMeta 11.1%
- 4 GPT-6.2 SolOpenAI 11.1%
- 7 Claude Opus 5.5Anthropic 22.2%
- 7 Muse EmberMeta 22.2%
- 7 GPT-6 LunaOpenAI 22.2%
- 10 Gemini 4 HydrogenGoogle 33.3%
- 11 DeepSeek-V4.1-Flash-1017DeepSeek 44.4%
- 11 Mercury 3Inception 44.4%
- 13 Qwen4-MaxAlibaba 55.6%
- 13 Particle-1Particle AI 55.6%
- 15 Gemini 3.5 Flash-LiteGoogle 66.7%
- 16 Claude Haiku 5.5Anthropic 77.8%
- 16 Gemini 3.1 Pro PreviewGoogle 77.8%
- 16 Gemini 3.8 FlashGoogle 77.8%
3 models had none: GPT-6.1 Sol, GPT-6.1 Terra, GPT-6.2 Astra.