01·Leaderboard
Every score below is at the effort you pick — not the press-release number.
Public leaderboards quote each model at maximum reasoning effort. Nobody runs xhigh for everything. Choose a tier and see what the models are actually worth there — models that don't support the tier resolve to their nearest one (shown per row).
ProprietaryOpen weights
↑ better · ← cheaperCost per typical task, log scale (1K in + 600 out + reasoning tokens)
| # | Model | Effort used | Index | GPQA | AIME | SWE-bench | Cost / task | $/M in · out | tok/s | Ctx |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Claude Fable 5 Anthropic | Low | 70.7 | 73.8 | 67.2 | 70.5 | $0.080 | $10 · $50 | 65 | 1M |
| 2 | GPT-5.5 OpenAI | Low | 69.4 | 73.0 | 67.9 | 67.1 | $0.012 | $1.25 · $10 | 140 | 400K |
| 3 | Claude Opus 4.8 Anthropic | Low | 68.6 | 72.2 | 65.1 | 68.8 | $0.037 | $5 · $25 | 55 | 1M |
| 4 | Gemini 3 Pro Google | Low | 63.7 | 68.3 | 56.3 | 61.6 | $0.016 | $2 · $12 | 130 | 1M |
| 5 | Claude Sonnet 5 Anthropic | Low | 62.4 | 65.6 | 54.9 | 62.7 | $0.019 | $3 · $15 | 80 | 1M |
| 6 | Grok 4.1 xAI | Low | 61.6 | 65.3 | 54.5 | 59.2 | $0.022 | $3 · $15 | 90 | 256K |
| 7 | GLM-5.2 Zhipu AI | Low | 61.3 | 63.8 | 53.4 | 61.6 | $0.0032 | $0.60 · $2.20 | 95 | 200K |
| 8 | DeepSeek-V3.2 DeepSeek | Low | 60.9 | 64.8 | 56.1 | 59.4 | $0.0007 | $0.28 · $0.42 | 60 | 128K |
| 9 | Qwen3-235B-A22B Alibaba | Low | 58.5 | 63.3 | 53.6 | 56.1 | $0.0038 | $0.70 · $2.80 | 70 | 128K |
| 10 | Kimi K2 Thinking Moonshot AI | Low | 54.1 | 57.2 | 42.6 | 52.4 | $0.0043 | $0.60 · $2.50 | 50 | 256K |
| 11 | Llama 4 Maverick Meta | Standard | 52.4 | 70.0 | 40.0 | 44.0 | $0.0008 | $0.27 · $0.85 | 160 | 1M |
Index = weighted composite of 8 benchmarks (weights on /benchmarks) · cost/task assumes 1K in + 600 out + the tier's typical reasoning tokens · preview data.