Frontier models, benchmarked continuously.
Quality, cost and speed for the leading AI models — refreshed on a live cycle · last updated 2026-11-22.
Intelligence, coding & agentic indexes
Composite indexes by Artificial Analysis · price per 1M tokens (input / output)
| # | Model | Intelligence | Coding | Agentic | $ / 1M in | $ / 1M out |
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5.2 (Max, Default Fallback) | 64.6 | — | — | $16.00 | $80.00 |
| 2 | GPT-6.2 Astra (Max) | 62.1 | — | — | $4.00 | $20.00 |
| 3 | Gemini 4 Neon (Xhigh) | 61.4 | — | — | $5.40 | $27.00 |
| 4 | Gemini 4 Carbon (High) | 60 | — | — | $2.20 | $11.00 |
| 5 | Claude Opus 5.5 (Max, Default Fallback) | 59.1 | — | — | $4.40 | $22.00 |
| 6 | GPT-6.2 Sol (Max) | 57.3 | — | — | $1.00 | $5.00 |
| 7 | GPT-6 Astra (Max) | 56.9 | — | — | $10.00 | $50.00 |
| 8 | GPT-6.1 Sol (Max) | 54.6 | — | — | $2.00 | $10.00 |
| 9 | Claude Sonnet 5.5 (Max, Default Fallback) | 54.5 | — | — | $2.20 | $11.00 |
| 10 | Gemini 4 Hydrogen (Xhigh) | 53.9 | — | — | $1.65 | $8.25 |
| 11 | GPT-6.1 Terra (Max) | 51.3 | — | — | $1.20 | $6.00 |
| 12 | Gemini 4 Argon (High) | 50.4 | — | — | $2.20 | $11.00 |
| 13 | Claude Fable 5.1 (Xhigh, Default Fallback) | 50 | — | — | $11.00 | $55.00 |
| 14 | Kimi K3.1 Pro | 46.5 | — | — | $4.00 | $20.00 |
| 15 | Claude Opus 5 (Max) | 44.2 | — | — | $5.50 | $27.50 |
| 16 | GPT-6 Sol (Max) | 44.2 | — | — | $2.00 | $10.00 |
| 17 | Muse Wonder (Max) | 43.6 | — | — | $4.00 | $20.00 |
| 18 | Qwen4 Max | 42.9 | — | — | $1.50 | $4.50 |
| 19 | Muse Spark 1.3 (Max) | 36.6 | — | — | $1.25 | $4.25 |
| 20 | Claude Haiku 5.5 (Max) | 36.5 | — | — | $0.110 | $0.550 |
| 21 | Muse Ember (Max) | 36 | — | — | $0.100 | $0.500 |
| 22 | Grok 4.7 (Xhigh) | 35.5 | — | — | $1.60 | $4.80 |
| 23 | Tencent Hy4 | 34.6 | — | — | $0.830 | $2.50 |
| 24 | Qwen3.8 Max (0902) | 33.7 | — | — | $1.15 | $3.45 |
| 25 | DeepSeek V4.1 Flash 1017 (Max) | 32.6 | — | — | $0.100 | $0.600 |
| 26 | GPT-5.6 Terra (Max) | 32.3 | — | — | $4.00 | $24.00 |
| 27 | GLM-5.3 (Max) | 31.7 | — | — | $1.40 | $4.40 |
| 28 | Grok 4.6 (Xhigh) | 31 | — | — | $1.60 | $4.80 |
| 29 | Gemini 3.8 Flash (High) | 30.8 | — | — | $0.830 | $4.12 |
| 30 | Inkling 2 Small | 30.6 | — | — | $0.290 | $0.950 |
Reasoning, agents & search — re-run on live models
GPQA Diamond
Graduate-level science reasoning
| # | Model | Score | $ / task |
|---|---|---|---|
| 1 | Gemini 4 Argon | 96.5% | $0.095 |
| 2 | Gemini 3.8 Flash | 95.6% | $0.068 |
| 3 | GPT-6 Astra Pro | 95.5% | $0.333 |
| 4 | Claude Fable 5.2 | 94.9% | $0.084 |
| 5 | Gemini 3.1 Pro Preview | 94.5% | $0.202 |
| 6 | GPT-6 Astra | 94.4% | $0.113 |
| 7 | GPT-6.1 Sol | 94.4% | $0.021 |
| 8 | GPT-5.6 Sol Pro | 94.1% | $0.151 |
| 9 | GPT-5.5 | 93.9% | $0.313 |
| 10 | Gemini 3.7 Flash | 93.9% | $0.029 |
| 11 | Kimi K3.1 Pro | 93.6% | $0.077 |
| 12 | Grok 4.6 | 92.8% | $0.123 |
| 13 | Claude Sonnet 5.5 | 92.8% | $0.025 |
| 14 | Gemini 3.6 Flash | 92.8% | $0.044 |
| 15 | GPT-5.6 Sol | 92.4% | $0.039 |
No single model wins every benchmark.
The leader changes by task: one model tops reasoning, another tops coding, a third wins agentic search. That is exactly why MultipleChat exists — put the leaders side by side, let them draft, challenge and verify each other, and keep the best answer.
Quality, cost and speed together
A benchmark score without a price tag is marketing. Every leaderboard here pairs quality with cost per task or per million tokens.
Live & open
Refreshed continuously and downloadable as JSON. Our own quarterly comparison study lives at the MultipleChat AI Benchmark.
Attribution
Source: OpenRouter (openrouter.ai/rankings) incl. Artificial Analysis indexes, as of 2026-11-22T20:11:44Z. Licensed under CC BY 4.0.