Compare models
Pick two models to see them side by side: the index, each capability and benchmark, and what a task costs and how long it takes.
Neighbours on the leaderboard
Each model against the one ranked just below it.
- GPT-6 Astra82.2%Gemini 3.8 Flash80.6%GPT-6 Astra +1.6 points
- Gemini 3.8 Flash80.6%Claude Fable 5.176.2%Gemini 3.8 Flash +4.4 points
- Claude Fable 5.176.2%Qwen3.8-Max73.8%Claude Fable 5.1 +2.4 points
- Qwen3.8-Max73.8%Kimi K371.9%Qwen3.8-Max +1.9 points
- Kimi K371.9%Qwen3.8-27B68.5%Kimi K3 +3.4 points
- Qwen3.8-27B68.5%Qwen3.5-35B-A3B65.7%Qwen3.8-27B +2.7 points
- Qwen3.5-35B-A3B65.7%Nemotron 3 Nano Omni 30B-A3B60.3%Qwen3.5-35B-A3B +5.4 points
- Nemotron 3 Nano Omni 30B-A3B60.3%Molmo 2 8B57.6%Nemotron 3 Nano Omni 30B-A3B +2.7 points
Against the leader
Every other model against GPT-6 Astra, ranked first.
- GPT-6 Astra82.2%Gemini 3.8 Flash80.6%GPT-6 Astra +1.6 points
- GPT-6 Astra82.2%Claude Fable 5.176.2%GPT-6 Astra +6.0 points
- GPT-6 Astra82.2%Qwen3.8-Max73.8%GPT-6 Astra +8.3 points
- GPT-6 Astra82.2%Kimi K371.9%GPT-6 Astra +10.3 points
- GPT-6 Astra82.2%Qwen3.8-27B68.5%GPT-6 Astra +13.7 points
- GPT-6 Astra82.2%Qwen3.5-35B-A3B65.7%GPT-6 Astra +16.4 points
- GPT-6 Astra82.2%Nemotron 3 Nano Omni 30B-A3B60.3%GPT-6 Astra +21.9 points
- GPT-6 Astra82.2%Molmo 2 8B57.6%GPT-6 Astra +24.5 points