Rock8Router
Log in(opens in a new tab)

Benchmarks

Artificial Analysis's published measurements, shown unchanged. The list is reordered to lead with labs based in the EU, then the rest of Europe — and every row keeps the rank the source gave it, so you can see where a model really places.

Figures published by Artificial Analysis (opens in a new tab), whose measurements these are — not ours. Last read 19 October 2026.

Leading European models

How the European labs compare with each other. Every other scored model is in the table below, and in the plot at the foot of this section.

Intelligence

The source's composite of every evaluation it runs. Higher is better.

  1. Mistral Large 4 Preview38.4
  2. Quasar 438B (Max, Based on GLM-5.2)26.7
  3. Mistral Medium 3.514.2
  4. Mistral Small 4 (Reasoning)11.3
  5. Mistral Large 39.3
  6. Mistral Medium 39.0
  7. Mistral Small 4 (Non-reasoning)9.0
  8. Magistral Small 1.28.6

Measured in 0–100. Bars are scaled to the largest figure shown, not to the full range.

Coding

The source's composite of its coding evaluations. Higher is better.

  1. Quasar 438B (Max, Based on GLM-5.2)61.2
  2. Mistral Medium 3.546.9
  3. Mistral Small 4 (Reasoning)26.6
  4. Mistral Large 320.1
  5. Magistral Small 1.214.7

Measured in 0–100. Bars are scaled to the largest figure shown, not to the full range.

Time to first token

How long before anything comes back at all. Lower is better — the bars are lengths, not scores.

  1. Mistral Large 4 Preview0.76s
  2. Quasar 438B (Max, Based on GLM-5.2)0.56s
  3. Mistral Medium 3.50.65s
  4. Mistral Small 4 (Reasoning)0.62s
  5. Mistral Small 4 (Non-reasoning)0.55s
  6. Magistral Small 1.20.80s

Measured in seconds. Bars are scaled to the largest figure shown, not to the full range.

Output speed

Tokens a second once the answer starts. Higher is better.

  1. Mistral Large 4 Preview62
  2. Quasar 438B (Max, Based on GLM-5.2)163
  3. Mistral Medium 3.5155
  4. Mistral Small 4 (Reasoning)187
  5. Mistral Small 4 (Non-reasoning)172
  6. Magistral Small 1.276

Measured in tok/s. Bars are scaled to the largest figure shown, not to the full range.

Against the whole field

Capability against price

Every scored model the source publishes, plotted by what it costs and how well it scores. Up and to the left is the good corner. The price axis is logarithmic — each step is ten times the last.

Intelligence index against blended price per million tokensA scatter plot of 426 models. The horizontal axis is price on a logarithmic scale; the vertical axis is the intelligence index. The8 models named on the chart are listed with their figures in the table below.0163248$0.01$0.10$1$10$100$1,000Blended price per million tokens (3:1 input to output)Intelligence indexMistral Small 3.1 — 7.1, $0.09 / 1MMistral Small 3.2 — 7.0, $0.11 / 1MMistral Large 2 (Jul '24) — 6.8, $3.00 / 1MMistral Small 3 — 6.7, $0.06 / 1MMinistral 3 14B — 6.0, $0.20 / 1MMistral Large (Feb '24) — 5.8, $6.00 / 1MMinistral 3 8B — 5.5, $0.15 / 1MMixtral 8x7B Instruct — 5.1, $0.51 / 1MMistral 7B Instruct — 5.0, $0.16 / 1MMinistral 3 3B — 4.8, $0.10 / 1MApertus 8B Instruct — 4.8, $0.13 / 1MClaude Opus 5.5 (Max, Default Fallback) — 57.6, $8.00 / 1MClaude Opus 5.5 (Xhigh, Default Fallback) — 56.0, $8.00 / 1MClaude Sonnet 5.5 (Max, Default Fallback) — 56.0, $4.00 / 1MClaude Opus 5.5 (High, Default Fallback) — 53.6, $8.00 / 1MClaude Fable 5.1 (Max, Default Fallback) — 53.4, $20.00 / 1MClaude Fable 5.1 (Xhigh, Default Fallback) — 53.2, $20.00 / 1MGPT-6 Astra (Max) — 52.7, $20.00 / 1MGemini 4 Argon (High) — 52.6, $4.00 / 1MGPT-6 Astra (Xhigh) — 52.4, $20.00 / 1MMuse Wonder (Max) — 52.4, $8.00 / 1MClaude Sonnet 5.5 (Xhigh, Default Fallback) — 51.9, $4.00 / 1MGPT-6.1 Sol (Max) — 51.8, $4.00 / 1MClaude Fable 5.1 (High, Default Fallback) — 51.2, $20.00 / 1MClaude Opus 5.5 (Medium, Default Fallback) — 51.2, $8.00 / 1MGPT-6.1 Sol (Xhigh) — 51.0, $4.00 / 1MGPT-6 Astra (High) — 50.9, $20.00 / 1MClaude Opus 5 (Max) — 50.8, $10.00 / 1MGPT-6.1 Sol (High) — 50.2, $4.00 / 1MClaude Opus 5 (Xhigh) — 49.7, $10.00 / 1MClaude Fable 5 (Max, Opus 4.8 Fallback) — 49.6, $20.00 / 1MGPT-6 Astra (Medium) — 49.6, $20.00 / 1MClaude Fable 5.1 (Medium, Default Fallback) — 48.9, $20.00 / 1MClaude Opus 5 (High) — 48.1, $10.00 / 1MMuse Spark 1.3 (Max) — 48.1, $2.00 / 1MGPT-6.1 Sol (Medium) — 47.8, $4.00 / 1MGPT-6 Sol (Max) — 47.6, $4.00 / 1MGPT-5.6 Sol (Max) — 47.0, $8.00 / 1MClaude Fable 5.1 (Low, Default Fallback) — 46.8, $20.00 / 1MClaude Sonnet 5.5 (High, Default Fallback) — 46.8, $4.00 / 1MGrok 4.7 (Xhigh) — 46.4, $3.00 / 1MGrok 4.7 (High) — 46.3, $3.00 / 1MMiMo-V2.6-Pro — 46.3, $0.54 / 1MGPT-6 Astra (Low) — 45.8, $20.00 / 1MQwen3.8 Max (0902) — 45.4, $3.00 / 1MMuse Spark 1.3 (Xhigh) — 45.1, $2.00 / 1MClaude Opus 5 (Medium) — 44.8, $10.00 / 1MGLM-5.3 (Max) — 44.8, $2.15 / 1MGrok 4.6 (High) — 44.3, $3.00 / 1MGPT-6 Sol (Xhigh) — 44.2, $4.00 / 1MGrok 4.6 (Xhigh) — 44.2, $3.00 / 1MGPT-5.6 Sol (Xhigh) — 44.0, $8.00 / 1MStep 5 Preview — 43.7, $1.43 / 1MKimi K3 (Max) — 43.6, $6.00 / 1MClaude Haiku 5.5 (Max) — 43.4, $0.20 / 1MGrok 4.6 (Medium) — 42.8, $3.00 / 1MGPT-6 Sol (High) — 42.4, $4.00 / 1MClaude Opus 5.5 (Low, Default Fallback) — 42.3, $8.00 / 1MGPT-5.6 Sol (High) — 42.3, $8.00 / 1MGrok 4.7 (Low) — 42.2, $3.00 / 1MGPT-5.6 Terra (Max) — 42.1, $4.50 / 1MGPT-6.1 Sol (Low) — 42.1, $4.00 / 1MClaude Opus 4.8 (Max) — 41.8, $10.00 / 1MGLM-5.3-Flash — 41.8, $0.24 / 1MClaude Haiku 5.5 (Xhigh) — 41.2, $0.20 / 1MLing 3.1 Flash — 41.1, $0.45 / 1MGemini 3.8 Flash (High) — 40.9, $1.50 / 1MClaude Sonnet 5.5 (Medium, Default Fallback) — 40.8, $4.00 / 1MClaude Opus 4.7 (Max) — 40.7, $10.00 / 1MQwen3.8 Max — 40.2, $3.00 / 1MQwen3.8 2.4T A95B — 39.9, $3.00 / 1MGemini 3.8 Flash (Medium) — 39.8, $1.50 / 1MGPT-6 Sol (Medium) — 39.8, $4.00 / 1MQwen3.8-Flash-Next — 39.8, $0.23 / 1MGemini 3.7 Flash (Medium) — 39.6, $1.50 / 1MMuse Spark 1.2 (Xhigh) — 39.6, $2.00 / 1MDeepSeek V4.1 Flash (Max) — 39.5, $0.53 / 1MClaude Opus 5 (Low) — 39.4, $10.00 / 1MGPT-5.6 Sol (Medium) — 39.2, $8.00 / 1MGemini 3.7 Flash (High) — 39.1, $1.50 / 1MGPT-5.4 (Xhigh) — 39.0, $5.63 / 1MMuse Spark Open (Xhigh) — 38.9, $2.00 / 1MGrok 4.5 (High) — 38.8, $3.00 / 1MGPT-5.5 (Xhigh) — 38.4, $11.25 / 1MClaude Sonnet 5 (Max) — 38.2, $4.00 / 1MGPT-6 Luna (Max) — 38.1, $0.20 / 1MGPT-5.6 Terra (Xhigh) — 38.0, $4.50 / 1MMiMo-V2.6-Flash — 37.9, $0.17 / 1MClaude Haiku 5.5 (High) — 37.8, $0.20 / 1MGPT-5.6 Luna (Max) — 37.3, $0.45 / 1MGPT-5.5 (High) — 37.0, $11.25 / 1MGemini 3.7 Flash (Low) — 36.9, $1.50 / 1MDeepSeek V4 Pro 0813 (Max) — 36.0, $1.98 / 1MClaude Sonnet 5.5 (Low, Default Fallback) — 35.9, $4.00 / 1MAgnes 2.5 Pro Beta — 35.2, $0.15 / 1MGrok 4.6 (Low) — 35.1, $3.00 / 1MDeepSeek V4 Flash Vision (Max) — 34.8, $0.66 / 1MGPT-5.6 Luna (Xhigh) — 34.6, $0.45 / 1MGPT-6 Luna (Xhigh) — 34.6, $0.20 / 1MClaude Haiku 5.5 (Medium) — 34.5, $0.20 / 1MClaude Sonnet 5 (Xhigh) — 34.4, $4.00 / 1MDeepSeek V4 Flash 0731 (Max) — 34.3, $0.17 / 1MGLM-5.3 (Low) — 34.3, $2.15 / 1MGPT-5.6 Terra (High) — 34.2, $4.50 / 1MGPT-6 Sol (Low) — 34.2, $4.00 / 1MGemini 3.6 Flash (High) — 34.0, $1.50 / 1MGPT-5.5 (Medium) — 33.8, $11.25 / 1MGLM-5.2 (Max) — 33.7, $2.15 / 1MMuse Spark 1.1 (Xhigh) — 33.7, $2.00 / 1MQwen3.8 27B (Xhigh) — 33.7, $1.13 / 1MGemini 3.5 Flash (Medium) — 33.6, $3.38 / 1MGemini 3.8 Flash (Low) — 33.5, $1.50 / 1MGPT-5.6 Sol (Low) — 33.5, $8.00 / 1MGPT-6 Luna (High) — 32.9, $0.20 / 1MGemini 3.5 Flash (High) — 32.6, $3.38 / 1MGPT-5.3 Codex (Xhigh) — 32.5, $4.81 / 1MGPT-5.6 Luna (High) — 32.1, $0.45 / 1MClaude Opus 4.6 (Max) — 31.9, $10.00 / 1MClaude Sonnet 5 (High) — 31.7, $4.00 / 1MClaude Opus 4.7 (Non-reasoning, High) — 30.9, $10.00 / 1MGPT-5.5 (Low) — 30.7, $11.25 / 1MDeepSeek V4 Pro 0424 (Max) — 30.4, $0.54 / 1MGPT-5.2 (Xhigh) — 30.4, $4.81 / 1MBeam 1.0 — 30.3, $1.05 / 1MClaude Sonnet 4.6 (Max) — 30.1, $6.00 / 1MDeepSeek V4 Pro 0424 (High) — 30.1, $0.54 / 1MGPT-5.6 Terra (Medium) — 30.1, $4.50 / 1MKimi K3 (Low) — 30.1, $6.00 / 1MGPT-6 Luna (Medium) — 29.9, $0.20 / 1MGemini 3.1 Pro Preview — 29.7, $4.50 / 1MClaude Haiku 5.5 (Low) — 29.5, $0.20 / 1MQwen3.7 Max — 29.5, $3.75 / 1MMiniMax-M3 — 29.2, $0.53 / 1MClaude Opus 4.5 (Reasoning) — 29.1, $10.00 / 1MGPT-5.2 Codex (Xhigh) — 28.5, $4.81 / 1MGPT-6 Sol (Non-reasoning) — 28.5, $4.00 / 1MGPT-5.6 Sol (Non-reasoning) — 28.3, $8.00 / 1MClaude Sonnet 5 (Medium) — 28.1, $4.00 / 1MGemini 3 Pro Preview (High) — 28.0, $4.50 / 1MGLM-5 (Reasoning) — 27.9, $1.55 / 1MGPT-5.4 (Low) — 27.6, $5.63 / 1MQwen3.8 27B (Medium) — 27.6, $1.13 / 1MGPT-5.6 Terra (Low) — 27.5, $4.50 / 1MGrok Build 0.1 0616 — 27.2, $1.25 / 1MKimi K2.6 (Reasoning) — 27.0, $1.71 / 1MQwen3.6 Plus — 27.0, $1.13 / 1MSolar Pro 4 — 26.6, $0.53 / 1MGPT-5.2 (Medium) — 26.5, $4.81 / 1MNemotron 3.5 Super VL — 26.5, $0.31 / 1MApodex 1.1 — 26.4, $0.97 / 1MClaude Opus 4.6 (Non-reasoning, High) — 26.4, $10.00 / 1MGemini 3 Flash Preview (Reasoning) — 26.3, $1.13 / 1MQwen3.8 27B (Low) — 26.2, $1.13 / 1MGLM-5.1 (Reasoning) — 26.1, $1.98 / 1MGPT-5.5 Instant (June 2026) — 26.0, $11.25 / 1MMiMo-V2.5-Pro (Reasoning) — 26.0, $0.54 / 1MKimi K2.7 Code — 25.8, $1.71 / 1MGrok 4.20 0309 v2 (Reasoning) — 25.7, $1.56 / 1MInkling Small — 25.7, $0.53 / 1MHy3 — 25.3, $0.24 / 1MGrok 4.20 0309 (Reasoning) — 25.2, $3.00 / 1MQwen3.7 Plus — 25.2, $0.70 / 1MGPT-5.6 Luna (Medium) — 25.0, $0.45 / 1MInkling (Xhigh) — 25.0, $1.76 / 1MGPT-5 Codex (High) — 24.9, $3.44 / 1MGrok 4.3 (High) — 24.9, $1.56 / 1MGrok 4.3 (Medium) — 24.8, $1.56 / 1MClaude Sonnet 4.6 (Non-reasoning, High) — 24.7, $6.00 / 1MDeepSeek V4.1 Flash (Non-reasoning) — 24.7, $0.53 / 1MGPT-5.1 (High) — 24.7, $3.44 / 1MDeepSeek V4 Flash 0420 (High) — 24.4, $0.14 / 1MClaude Sonnet 5 (Low) — 24.3, $4.00 / 1MGrok 4.3 (Low) — 24.3, $1.56 / 1MDeepSeek V4 Flash 0420 (Max) — 24.2, $0.14 / 1MGLM-5.1 (Non-reasoning) — 24.2, $2.13 / 1MGPT-5.4 mini (Xhigh) — 24.1, $1.69 / 1MSolar Mini 4 — 24.1, $0.17 / 1MGemini 3.5 Flash (Minimal) — 23.8, $3.38 / 1MClaude Opus 4.5 (Non-reasoning) — 23.7, $10.00 / 1MGPT-5.1 Codex (High) — 23.7, $3.44 / 1MKimi K2.6 (Non-reasoning) — 23.6, $1.71 / 1MKimi K2.5 (Reasoning) — 23.5, $1.14 / 1MClaude Sonnet 4.6 (Non-reasoning, Low) — 23.3, $6.00 / 1MClaude Sonnet 5 (Non-reasoning, High) — 23.2, $4.00 / 1MGPT-5.5 (Non-reasoning) — 23.2, $11.25 / 1MGPT-5 (High) — 23.0, $3.44 / 1MGPT-5 (Medium) — 22.9, $3.44 / 1MNemotron 3 Ultra 550B A55B (Reasoning) — 22.9, $1.13 / 1MQwen3.5 27B (Reasoning) — 22.9, $0.82 / 1MClaude 4.1 Opus (Reasoning) — 22.8, $30.00 / 1MMiniMax-M2.5 — 22.8, $0.53 / 1MMiniMax-M2.7 — 22.8, $0.53 / 1MGPT-5.5 Instant (May 2026) — 22.7, $11.25 / 1MHy3-preview (Reasoning) — 22.7, $0.10 / 1MLing-3.0-flash-Fin — 22.6, $0.11 / 1MGrok 4 — 22.5, $6.00 / 1MGLM-5.2 (Non-reasoning) — 22.4, $2.15 / 1MGemini 3 Pro Preview (Low) — 22.3, $4.50 / 1MGemini 3.5 Flash-Lite — 22.2, $0.85 / 1MGLM-4.7 (Reasoning) — 22.2, $1.00 / 1MKimi K2 Thinking — 22.0, $1.07 / 1Mo3-pro — 21.9, $35.00 / 1MGLM-5 (Non-reasoning) — 21.8, $1.55 / 1MMiMo-V2.5 — 21.8, $0.17 / 1MDeepSeek V3.2 (Reasoning) — 21.5, $0.32 / 1MGPT-6 Luna (Low) — 21.5, $0.20 / 1MQwen3.5 397B A17B (Non-reasoning) — 21.4, $1.35 / 1MQwen3.6 27B (Reasoning) — 21.4, $1.35 / 1MGPT-5.6 Luna (Low) — 21.0, $0.45 / 1MMiniMax-M2.1 — 20.9, $0.53 / 1MDeepSeek V4 Pro 0424 (Non-reasoning) — 20.8, $0.54 / 1MGPT-5 (Low) — 20.8, $3.44 / 1MGPT-5.6 Terra (Non-reasoning) — 20.8, $4.50 / 1MMiMo-V2-Flash (Reasoning) — 20.8, $0.15 / 1MClaude 4.5 Sonnet (Reasoning) — 20.7, $6.00 / 1MGPT-5.4 nano (Xhigh) — 20.7, $0.46 / 1MClaude 4 Opus (Reasoning) — 20.6, $30.00 / 1MGPT-5 mini (Medium) — 20.6, $0.69 / 1MDeepSeek V4 Pro 0813 (Non-reasoning) — 20.4, $1.98 / 1MGPT-5.1 Codex mini (High) — 20.4, $0.69 / 1MQwen3.5 Omni Plus — 20.4, $1.50 / 1Mo3 — 20.2, $3.50 / 1MQwen3.8 27B (Non-reasoning) — 20.2, $1.13 / 1MLing 3.0 Flash — 20.1, $0.11 / 1MGPT-5.4 nano (Medium) — 20.0, $0.46 / 1MQwen3.6 27B (Non-reasoning) — 19.8, $1.35 / 1MGPT-5.4 mini (Medium) — 19.7, $1.69 / 1MStep 3.7 Flash — 19.5, $0.44 / 1MKimi K2.5 (Non-reasoning) — 19.4, $1.20 / 1MQwen3.5 27B (Non-reasoning) — 19.4, $0.82 / 1MClaude 4.5 Sonnet (Non-reasoning) — 19.3, $6.00 / 1MQwen3.5 35B A3B (Reasoning) — 19.3, $0.69 / 1MLongCat 2.0 — 19.1, $1.30 / 1MDeepSeek V4 Flash 0420 (Non-reasoning) — 18.9, $0.12 / 1MClaude 4.1 Opus (Non-reasoning) — 18.6, $30.00 / 1MMiniMax-M2 — 18.6, $0.53 / 1MGLM-4.6 (Reasoning) — 18.5, $0.96 / 1MGPT-6 Luna (Non-reasoning) — 18.5, $0.20 / 1MQwen3.5 397B A17B (Reasoning) — 18.4, $1.35 / 1MMiMo-V2.5-Pro (Non-reasoning) — 18.3, $0.54 / 1MGPT-5.4 (Non-reasoning) — 18.2, $5.63 / 1MQwen3.6 35B A3B (Reasoning) — 18.2, $0.84 / 1MGemini 3 Flash Preview (Non-reasoning) — 17.9, $1.13 / 1MGrok 4 Fast (Reasoning) — 17.9, $0.28 / 1MQwen3.5 122B A10B (Non-reasoning) — 17.7, $1.10 / 1MMuse Glimmer (High) — 17.5, $0.53 / 1MGLM-4.7 (Non-reasoning) — 17.4, $1.00 / 1MGPT-5.2 (Non-reasoning) — 17.0, $4.81 / 1MHy3-preview (Non-reasoning) — 17.0, $0.10 / 1MLing-2.6-1T — 17.0, $0.85 / 1MStep 3.5 Flash 2603 — 17.0, $0.15 / 1MClaude 4.5 Haiku (Reasoning) — 16.9, $2.00 / 1MGPT-5 mini (High) — 16.8, $0.69 / 1MGemma 4 26B A4B (Reasoning) — 16.7, $0.17 / 1Mo4-mini (High) — 16.7, $1.93 / 1MClaude 4 Opus (Non-reasoning) — 16.6, $30.00 / 1MDeepSeek V3.2 Exp (Reasoning) — 16.6, $0.32 / 1MRing-2.6-1T — 16.6, $0.85 / 1MStep 3.5 Flash — 16.6, $0.15 / 1MGemini 2.5 Pro — 16.1, $3.44 / 1MDeepSeek V3.2 (Non-reasoning) — 16.0, $0.32 / 1MGemini 3.1 Flash-Lite — 15.6, $0.56 / 1MQwen3.5 122B A10B (Reasoning) — 15.6, $1.10 / 1MGPT-5.6 Luna (Non-reasoning) — 15.5, $0.45 / 1MClaude 4.5 Haiku (Non-reasoning) — 15.4, $2.00 / 1MClaude 3.7 Sonnet (Non-reasoning) — 15.3, $6.00 / 1MKimi K2 0905 — 15.3, $1.07 / 1Mo1 — 15.2, $26.25 / 1MQwen3.6 35B A3B (Non-reasoning) — 15.2, $0.84 / 1MQwen3.5 35B A3B (Non-reasoning) — 15.1, $0.69 / 1MGLM-4.6 (Non-reasoning) — 14.9, $0.98 / 1MGLM-4.7-Flash (Reasoning) — 14.9, $0.15 / 1MDeepSeek V3.1 Terminus (Reasoning) — 14.8, $1.91 / 1MGrok 3 mini Reasoning (High) — 14.6, $0.35 / 1MGrok 4.20 0309 (Non-reasoning) — 14.6, $3.00 / 1MGemini 2.5 Pro Preview (May' 25) — 14.5, $3.44 / 1MGemma 4 12B (Reasoning) — 14.2, $0.15 / 1MGrok 4.20 0309 v2 (Non-reasoning) — 14.2, $1.56 / 1MNova 2.0 Pro Preview (Medium) — 14.2, $3.44 / 1MGrok 4.3 (Non-reasoning) — 14.0, $1.56 / 1MDeepSeek V3.1 Terminus (Non-reasoning) — 13.9, $0.45 / 1MDeepSeek V3.2 Exp (Non-reasoning) — 13.9, $0.32 / 1MGemma 4 31B (Non-reasoning) — 13.9, $0.22 / 1MDeepSeek V3.1 (Non-reasoning) — 13.7, $0.84 / 1MDeepSeek V3.1 (Reasoning) — 13.5, $0.85 / 1MNova 2.0 Lite (High) — 13.4, $0.85 / 1MGPT-5.1 (Non-reasoning) — 13.3, $3.44 / 1MQwen3.5 9B (Non-reasoning) — 13.3, $0.19 / 1MDeepSeek R1 0528 (May '25) — 13.1, $1.76 / 1MGemini 2.5 Flash (Reasoning) — 13.1, $0.85 / 1MGemma 4 26B A4B (Non-reasoning) — 13.1, $0.20 / 1MQwen3.5 4B (Reasoning) — 13.1, $0.06 / 1MGPT-5 nano (High) — 13.0, $0.14 / 1MNemotron 3.5 Lightning — 12.9, $0.10 / 1MNemotron 3 Super 120B A12B (Reasoning) — 12.8, $0.45 / 1MNova 2.0 Pro Preview (Low) — 12.8, $3.44 / 1MGPT-4.1 — 12.7, $3.50 / 1MGranite 4.2 30B — 12.7, $0.28 / 1MKimi K2 — 12.7, $1.00 / 1MQwen3 235B A22B 2507 (Reasoning) — 12.7, $0.75 / 1MGPT-5 nano (Medium) — 12.5, $0.14 / 1MNova 2.0 Lite (Medium) — 12.5, $0.85 / 1Mo3-mini — 12.5, $1.93 / 1MQwen3.5 Omni Flash — 12.5, $0.28 / 1Mo1-pro — 12.4, $262.50 / 1MMercury 2.5 — 12.3, $0.38 / 1MGrok 3 — 12.1, $6.00 / 1MSeed-OSS-36B-Instruct — 12.1, $0.30 / 1MQwen3 235B A22B 2507 Instruct — 12.0, $0.39 / 1MQwen3 Coder 480B A35B Instruct — 11.9, $0.68 / 1MNova 2.0 Lite (Low) — 11.8, $0.85 / 1MGPT-5.4 nano (Non-reasoning) — 11.7, $0.46 / 1MMiniMax M1 80k — 11.7, $0.96 / 1Mgpt-oss-120b (High) — 11.6, $0.24 / 1MDeepSeek R1 (Jan '25) — 11.4, $2.50 / 1MGPT-5 (Minimal) — 11.4, $3.44 / 1Mo1-preview — 11.4, $28.88 / 1MGLM-4.6V (Reasoning) — 11.2, $0.45 / 1MQwen3 Next 80B A3B (Reasoning) — 11.2, $0.41 / 1MQwen3.5 9B (Reasoning) — 11.2, $0.15 / 1MGLM-4.5-Air — 11.1, $0.37 / 1MGPT-5.4 mini (Non-reasoning) — 11.1, $1.69 / 1MGranite 4.2 8B — 11.1, $0.11 / 1MGrok 4 Fast (Non-reasoning) — 11.1, $0.28 / 1Mo3-mini (High) — 11.0, $1.93 / 1MMercury 2 — 10.8, $0.38 / 1MQwen3.5 4B (Non-reasoning) — 10.8, $0.06 / 1MTrinity Large Thinking — 10.8, $0.39 / 1MGLM-4.7-Flash (Non-reasoning) — 10.6, $0.15 / 1MGemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning) — 10.4, $0.17 / 1MNemotron 3 Nano Omni 30B A3B Reasoning — 10.3, $0.45 / 1MGPT-4.1 mini — 10.2, $0.70 / 1Mgpt-oss-120b (Low) — 10.2, $0.22 / 1Mgpt-oss-20b (Low) — 10.0, $0.12 / 1MNova 2.0 Pro Preview (Non-reasoning) — 10.0, $3.44 / 1MGemini 2.5 Flash (Non-reasoning) — 9.9, $0.85 / 1MGPT-5 mini (Minimal) — 9.9, $0.69 / 1MQwen3 VL 235B A22B Instruct — 9.9, $0.51 / 1MQwen3 30B A3B 2507 (Reasoning) — 9.8, $0.59 / 1MDeepSeek V3 0324 — 9.7, $0.41 / 1MQwen3 Coder 30B A3B Instruct — 9.6, $0.30 / 1MQwen3 Next 80B A3B Instruct — 9.6, $0.38 / 1MQwen3 VL 30B A3B (Reasoning) — 9.5, $0.50 / 1MGemma 4 12B (Non-reasoning) — 9.4, $0.15 / 1MGemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning) — 9.3, $0.17 / 1MLlama 4 Maverick — 9.3, $0.42 / 1MQwen3 Coder Next — 9.2, $0.48 / 1MGranite 4.2 3B — 9.1, $0.05 / 1Mgpt-oss-20b (High) — 9.0, $0.10 / 1MGemma 4 E4B (Reasoning) — 8.9, $0.04 / 1MNVIDIA Nemotron 3 Nano 30B A3B (Reasoning) — 8.9, $0.09 / 1MSarvam 105B (High) — 8.8, $0.07 / 1MClaude 3 Opus — 8.7, $30.00 / 1MNova 2.0 Lite (Non-reasoning) — 8.7, $0.85 / 1MDeepSeek V3 (Dec '24) — 8.5, $0.46 / 1MGemini 2.5 Flash-Lite (Reasoning) — 8.5, $0.17 / 1MGLM-4.6V (Non-reasoning) — 8.4, $0.45 / 1MGPT-4o (Nov '24) — 8.4, $4.38 / 1MClaude 3.5 Sonnet (Oct '24) — 7.9, $6.00 / 1MDeepSeek R1 Distill Llama 70B — 7.9, $0.80 / 1MQwen3 VL 30B A3B Instruct — 7.9, $0.41 / 1MGPT-4.1 nano — 7.8, $0.17 / 1MLing-flash-2.0 — 7.8, $0.25 / 1MQwen3 Omni 30B A3B (Reasoning) — 7.8, $0.43 / 1MSolar Pro 3 — 7.8, $0.26 / 1MGPT-4o (Aug '24) — 7.7, $4.38 / 1MLlama 3.3 Instruct 70B — 7.7, $0.71 / 1MQwen2.5 Instruct 72B — 7.7, $0.48 / 1MGLM-4.5V (Reasoning) — 7.6, $0.90 / 1MQwen3 30B A3B (Reasoning) — 7.6, $0.24 / 1MGemma 4 E4B (Non-reasoning) — 7.5, $0.04 / 1MHermes 4 - Llama-3.1 405B (Reasoning) — 7.5, $1.50 / 1MQwen3 30B A3B 2507 Instruct — 7.5, $0.25 / 1MHermes 4 - Llama-3.1 405B (Non-reasoning) — 7.4, $1.50 / 1MNVIDIA Nemotron Nano 9B V2 (Reasoning) — 7.4, $0.07 / 1MGPT-4o (May '24) — 7.3, $7.50 / 1MQwen3 32B (Non-reasoning) — 7.3, $0.26 / 1MClaude 3.5 Sonnet (June '24) — 7.2, $6.00 / 1MQwen3 32B (Reasoning) — 7.2, $0.32 / 1MGPT-5 nano (Minimal) — 7.1, $0.14 / 1MCommand A — 7.0, $4.38 / 1MGPT-4 Turbo — 7.0, $15.00 / 1MNova Pro — 7.0, $1.40 / 1MLlama 3.1 Instruct 8B — 6.9, $0.04 / 1MNVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning) — 6.8, $0.09 / 1MNVIDIA Nemotron Nano 9B V2 (Non-reasoning) — 6.8, $0.09 / 1MGemini 2.5 Flash-Lite (Non-reasoning) — 6.7, $0.17 / 1MGLM-4.5V (Non-reasoning) — 6.7, $0.90 / 1MGPT-4 — 6.7, $37.50 / 1MGPT-4o mini — 6.7, $0.26 / 1MNova Lite — 6.7, $0.10 / 1MQwen3 14B (Non-reasoning) — 6.7, $0.15 / 1MLlama 3.1 Instruct 70B — 6.6, $0.56 / 1MQwen3 30B A3B (Non-reasoning) — 6.6, $0.21 / 1MLlama 4 Scout — 6.5, $0.30 / 1MQwen2.5 Turbo — 6.4, $0.09 / 1MQwen3 14B (Reasoning) — 6.4, $0.15 / 1MReka Flash (Sep '24) — 6.4, $0.35 / 1MSolar Mini — 6.4, $0.15 / 1MCeleris-1 — 6.3, $0.33 / 1MPhi-4 Mini Instruct — 6.3, $0.13 / 1MHermes 3 - Llama-3.1 70B — 6.0, $0.70 / 1MJamba 1.5 Large — 6.0, $3.50 / 1MQwen3 8B (Non-reasoning) — 6.0, $0.20 / 1MQwen3 Omni 30B A3B Instruct — 6.0, $0.43 / 1MNova Micro — 5.9, $0.06 / 1MPhi-4 — 5.9, $0.22 / 1MNVIDIA Nemotron Nano 12B v2 VL (Non-reasoning) — 5.8, $0.30 / 1MPhi-4 Multimodal Instruct — 5.8, $0.14 / 1MLlama 2 Chat 7B — 5.7, $0.10 / 1MClaude 3 Haiku — 5.6, $0.50 / 1MReka Flash 3 — 5.6, $0.35 / 1MGPT-3.5 Turbo — 5.5, $0.75 / 1MLlama 3 Instruct 70B — 5.5, $1.18 / 1MLlama 3.2 Instruct 11B (Vision) — 5.4, $0.34 / 1MJamba 1.5 Mini — 5.2, $0.25 / 1MGranite 3.3 8B (Non-reasoning) — 4.9, $0.09 / 1MLlama 3 Instruct 8B — 4.8, $0.07 / 1MMistral Large 4 Previ…Mistral Large 4 Preview — 38.4, $2.06 / 1MQuasar 438B (Max, Bas…Quasar 438B (Max, Based on GLM-5.2) — 26.7, $0.90 / 1MMistral Medium 3.5Mistral Medium 3.5 — 14.2, $3.00 / 1MMistral Small 4 (Reas…Mistral Small 4 (Reasoning) — 11.3, $0.26 / 1MMistral Large 3Mistral Large 3 — 9.3, $0.75 / 1MMistral Medium 3Mistral Medium 3 — 9.0, $0.80 / 1MMistral Small 4 (Non-…Mistral Small 4 (Non-reasoning) — 9.0, $0.26 / 1MMagistral Small 1.2Magistral Small 1.2 — 8.6, $0.75 / 1M
Intelligence index and blended price for the models named on the chart
ModelIntelligence indexBlended price per million tokens
Mistral Large 4 Preview38.4$2.06
Quasar 438B (Max, Based on GLM-5.2)26.7$0.90
Mistral Medium 3.514.2$3.00
Mistral Small 4 (Reasoning)11.3$0.26
Mistral Large 39.3$0.75
Mistral Medium 39.0$0.80
Mistral Small 4 (Non-reasoning)9.0$0.26
Magistral Small 1.28.6$0.75

Every measurement

The source scores a reasoning model once per effort setting and publishes a row for each, so one model can appear several times.

Showing 25 of 704 measurements — 37 from European labs.

Benchmarked models, sortable by any column. Rank is the source's own ranking by intelligence index and does not change when the table is re-sorted.
1Mistral Large 4 PreviewFR · EUMistral38.4———$2.065620.76s
2Quasar 438B (Max, Based on GLM-5.2)ES · EUMultiverse Computing26.761.2——$0.901630.56s
3Mistral Medium 3.5FR · EUMistral14.246.9——$3.001550.65s
4Magistral Medium 1.2FR · EUMistral11.821.382.0—$0.00——
5HyperNova 60B 2605 (High, Based on gpt-oss-120b)ES · EUMultiverse Computing11.723.2——$0.00——
6Mistral Small 4 (Reasoning)FR · EUMistral11.326.6——$0.2621870.62s
7Mistral Large 3FR · EUMistral9.320.138.0—$0.75——
8Mistral Medium 3.1FR · EUMistral9.220.538.3—$0.00——
9Magistral Medium 1FR · EUMistral9.1—40.3—$0.00——
10Devstral MediumFR · EUMistral9.0—4.7—$0.00——
11Mistral Medium 3FR · EUMistral9.0—30.3—$0.80——
12Mistral Small 4 (Non-reasoning)FR · EUMistral9.0———$0.2621720.55s
13Devstral Small (May '25)FR · EUMistral8.7———$0.00——
14Devstral 2FR · EUMistral8.631.336.7—$0.00——
15Magistral Small 1.2FR · EUMistral8.614.780.3—$0.75760.80s
16Magistral Small 1FR · EUMistral8.2—41.3—$0.00——
17Devstral Small (Jul '25)FR · EUMistral7.6—29.3—$0.00——
18Mistral Large 2 (Nov '24)FR · EUMistral7.6—14.0—$0.00——
19Devstral Small 2FR · EUMistral7.529.334.3—$0.00——
20Mistral Small 3.1FR · EUMistral7.126.33.7—$0.094——
21Pixtral LargeFR · EUMistral7.1—2.3—$0.00——
22Mistral Small 3.2FR · EUMistral7.012.527.0—$0.106——
23Mistral Large 2 (Jul '24)FR · EUMistral6.8—0.0—$3.00——
24Mistral Small 3FR · EUMistral6.7—4.3—$0.058——
25Mistral SabaFR · EUMistral6.5———$0.00——
Back to top