Leaderboard
AI Model Leaderboard
The only board that ranks LLMs on all three axes that matter to a builder: how smart they are (intelligence), how widely used they are (real OpenRouter traffic), and your value — intelligence per dollar.
Ranked by the Artificial Analysis Intelligence Index. Value = intelligence per blended $/1M token. Popularity shown where the model has OpenRouter traffic.
| # | Model | |||
|---|---|---|---|---|
| 1 | Claude Opus 5.5 | 57.6 | #16 76% | 4.80 |
| 2 | Claude Sonnet 5.5 | 56.0 | #30 146% | 9.33 |
| 3 | Claude Fable 5.1 | 53.4 | #43 37% | 1.78 |
| 4 | GPT-6 Astra | 52.7 | #25 65% | 1.76 |
| 5 | Gemini 4 Argon | 52.6 | — | 8.77 |
| 6 | GPT-6.1 Sol | 51.8 | #31 112% | 8.63 |
| 7 | Claude Opus 5 | 50.8 | #27 11% | 3.39 |
| 8 | Claude Fable 5 | 49.6 | #73 — | 1.65 |
| 9 | Muse Spark 1.3 | 48.1 | — | 17.49 |
| 10 | GPT-6 Sol | 47.6 | #33 35% | 7.93 |
| 11 | Grok 4.7 | 46.4 | #49 105% | 11.60 |
| 12 | MiMo-V2.6-Pro | 46.3 | #28 11% | 70.96 |
| 13 | Qwen3.8 Max | 45.4 | #61 35% | 11.35 |
| 14 | GLM-5.3 | 44.8 | #10 12% | 15.45 |
| 15 | Grok 4.6 | 44.3 | #64 — | 11.07 |
| 16 | Step 5 Preview | 43.7 | #14 — | 23.62 |
| 17 | Claude Haiku 5.5 | 43.4 | #52 — | 144.67 |
| 18 | GPT-5.6 Sol | 42.3 | #19 2% | 3.52 |
| 19 | GPT-5.6 Terra | 42.1 | #47 100% | 6.01 |
| 20 | GLM-5.3-Flash | 41.8 | #3 6% | 128.62 |
| 21 | Claude Opus 4.8 | 41.8 | #62 — | 2.79 |
| 22 | Ling 3.1 Flash | 41.1 | #60 149% | 68.50 |
| 23 | Gemini 3.8 Flash | 40.9 | #15 21% | 18.18 |
| 24 | Qwen3.8 2.4T A95B | 39.9 | — | 9.97 |
| 25 | Qwen3.8-Flash-Next | 39.8 | — | 128.39 |
| 26 | Muse Spark 1.2 | 39.6 | — | 14.40 |
| 27 | Gemini 3.7 Flash | 39.1 | #40 42% | 17.38 |
| 28 | GPT-5.4 | 39.0 | — | 4.46 |
| 29 | Grok 4.5 | 38.8 | — | 9.70 |
| 30 | GPT-5.5 | 38.4 | — | 2.19 |
| 31 | Mistral Large 4 Preview | 38.4 | — | 13.86 |
| 32 | Claude Sonnet 5 | 38.2 | #24 29% | 6.37 |
| 33 | GPT-6 Luna | 38.1 | #9 1% | 127.00 |
| 34 | MiMo-V2.6-Flash | 37.9 | #7 2% | 180.48 |
| 35 | GPT-5.6 Luna | 37.3 | #6 54% | 53.29 |
| 36 | DeepSeek V4 Pro 0813 | 36.0 | — | 13.64 |
| 37 | DeepSeek V4 Flash Vision | 34.8 | — | 39.55 |
| 38 | Gemini 3.6 Flash | 34.0 | — | 15.11 |
| 39 | JT-4.1 Flash 236B A21B | 33.9 | — | — |
| 40 | Qwen3.8 27B | 33.7 | #46 56% | 19.26 |
| 41 | Muse Spark 1.1 | 33.7 | — | 12.25 |
| 42 | Gemini 3.5 Flash | 33.6 | — | 6.40 |
| 43 | Motif 3 | 33.6 | — | — |
| 44 | GPT-5.3 Codex | 32.5 | — | 4.13 |
| 45 | Claude Opus 4.6 | 31.9 | — | 2.13 |
| 46 | Claude Opus 4.7 | 30.9 | — | 2.06 |
| 47 | K2 Horizon 375B A23B | 30.5 | — | 3050.00 |
| 48 | GPT-5.2 | 30.4 | — | 3.86 |
| 49 | DeepSeek V4 Pro 0424 | 30.4 | — | 46.59 |
| 50 | Kimi K3 | 30.1 | #20 10% | 3.34 |
The board is a superset of the calculator catalog: brand-new models appear here with quality & popularity before the pricing catalog catches up. Models with a calculator match link to a head-to-head comparison.
Data: Artificial Analysis · OpenRouter
Know your margin before you build
Pick a model above, then model your real cost and price in the Calcaas calculator — free.