Benchmarks say one thing, people using the models say another. We pull both — plus real market pricing and measured speed — and score 250 models on one scale. 128 of them carry a machine score and a human score, which is where the disagreements show up.
28 of these models have a hands-on review on this site.
What feeds the index
live market pricing
benchmark indices
human preference votes
128 models scored by machines and humans
Benchmark indices cross-checked against human preference votes. The bar under each score splits it into the five things we weigh — same total, different shape means a different model for a different job.
Benchmarks, human votes, speed and real pricing on the same screen. The highlighted side wins that line.
Claude Opus 5.5
anthropic
88.7
intelligence · wins 10 lines
Claude Fable 5.1
anthropic
88.2
intelligence · wins 1 line
Our composite score
| 64.7 | Overall | 60.1 |
|---|---|---|
| 89 | Intelligence40% of the total | 88 |
| 19 | Price / perf20% of the total | 0 |
| 85 | Capabilities15% of the total | 85 |
| 45 | Speed10% of the total | 38 |
| 55 | Ecosystem15% of the total | 55 |
Measured intelligence
| 57.6 | Intelligence indexArtificial Analysis | 53.4 |
|---|---|---|
| — | Coding indexArtificial Analysis | 81.6 |
| — | Agentic indexArtificial Analysis | 57.9 |
| 1512 · 5k votes | Human preferenceLMArena Elo | 1511 · 12k votes |
Speed
| 92 t/s | Output throughputtokens/second | 64 t/s |
|---|---|---|
| 742.91 s | Time to first token | 256.22 s |
Price per million tokens
| $4.0 | Input | $10 |
|---|---|---|
| $20 | Output | $50 |
| $16 | Blended 3:1typical real-world mix | $40 |
Capabilities
| 1M | Context window | 1M |
|---|---|---|
| 128k | Max output | 128k |
| Yes | Reasoning | Yes |
| No | Open weights | No |
| text, image, file | Modalities | text, image, file |
| 1 | Providers serving it | 1 |
| — | Knowledge cutoff | — |
Filter it down to the models that fit your constraints, then open a row for the full measurement set.
250 of 250 models
| # | Model | intelligence & profile | Intel. | Elo | t/s | In / Out | Context | Actions |
|---|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 anthropic | 88.7 | 57.6 | 1512 | 92 | $4.0 / $20 | 1M | |
| 2 | Claude Fable 5.1 anthropic | 88.2 | 53.4 | 1511 | 64 | $10 / $50 | 1M | |
| 3 | anthropic | 86.0 | 50.8 | 1507 | 56 | $5.0 / $25 | 1M | |
| 4 | anthropic | 82.4 | 49.6 | 1492 | 61 | $10 / $50 | 1M | |
| 5 | Muse Spark 1.3 meta | 82.3 | 48.1 | 1490 | 146 | $1.3 / $4.3 | 1.0M | |
| 6 | Claude Opus 5.5 (batch) anthropic | 82.3 | 57.6 | — | — | $2.0 / $10 | 1M | |
| 7 | Claude Fable 5.1 (batch) anthropic | 81.5 | 53.4 | — | — | $5.0 / $25 | 1M | |
| 8 | Qwen3.8 Max (0902) qwen | 80.5 | 45.4 | 1482 | 37 | $2.0 / $6.0 | 1M | |
| 9 | Claude Sonnet 5.5 (batch) anthropic | 80.0 | 56.0 | — | — | $1.0 / $5.0 | 1M | |
| 10 | Claude Sonnet 5.5 anthropic | 80.0 | 56.0 | — | 129 | $2.0 / $10 | 1M | |
| 11 | GLM 5.3OPEN z-ai | 78.2 | 44.8 | 1471 | 71 | $1.4 / $4.4 | 1.0M | |
| 12 | Claude Opus 5 (batch) anthropic | 78.1 | 50.8 | — | — | $2.5 / $13 | 1M | |
| 13 | Qwen3.6 Max Preview qwen | 78.1 | — | 1447 | — | $1.0 / $6.2 | 262k | |
| 14 | Gemini 3.8 Flash google | 78.0 | 40.9 | 1497 | 237 | $0.75 / $3.8 | 1.0M | |
| 15 | Kimi K3OPEN moonshotai | 78.0 | 43.6 | 1476 | 43 | $0.67 / $14 | 1.0M | |
| 16 | MiMo-V2.6-ProOPEN xiaomi | 77.9 | 46.3 | 1491 | 41 | $0.43 / $0.87 | 1.1M | |
| 17 | GPT-6 Astra (batch) openai | 77.2 | 52.7 | — | — | $5.0 / $25 | 1.1M | |
| 18 | GPT-6 Astra openai | 77.1 | 52.7 | 1442 | 60 | $10 / $50 | 1.1M | |
| 19 | anthropic | 76.4 | 40.7 | 1490 | 46 | $5.0 / $25 | 1M | |
| 20 | GLM 5.3 FlashOPEN z-ai | 76.0 | 41.8 | 1470 | 52 | $0.15 / $0.50 | 1.0M | |
| 21 | GPT-6.1 Sol openai | 75.9 | 51.8 | 1446 | 54 | $2.0 / $10 | 1.1M | |
| 22 | Claude Fable 5 (batch) anthropic | 74.8 | 49.6 | — | — | $5.0 / $25 | 1M | |
| 23 | Gemini 3.7 Flash google | 74.7 | 39.6 | 1487 | 271 | $0.75 / $3.8 | 1.0M | |
| 24 | openai | 73.7 | 39.0 | 1469 | 105 | $2.5 / $15 | 1.1M | |
| 25 | anthropic | 73.7 | 41.8 | 1461 | 56 | $5.0 / $25 | 1M | |
| 26 | MiMo-V2.5OPEN xiaomi | 73.0 | — | 1428 | — | $0.14 / $0.28 | 1.1M | |
| 27 | GPT-5.6 Sol (batch) openai | 73.0 | 47.0 | — | — | $1.0 / $5.0 | 1.1M | |
| 28 | openai | 73.0 | 47.0 | — | 81 | $2.0 / $10 | 1.1M | |
| 29 | openai | 72.9 | 38.4 | 1471 | 92 | $5.0 / $30 | 1.1M | |
| 30 | x-ai | 72.4 | 44.3 | 1427 | 73 | $2.0 / $6.0 | 500k | |
| 31 | GLM 5.3 (batch)OPEN z-ai | 71.7 | 44.8 | — | — | $0.45 / $2.0 | 1.0M | |
| 32 | GLM 5.2OPEN z-ai | 70.6 | 33.7 | 1470 | 94 | $0.02 / $16 | 1.0M | |
| 33 | x-ai | 70.5 | 38.8 | 1448 | 59 | $2.0 / $6.0 | 500k | |
| 34 | Gemini 3.6 Flash google | 70.3 | 34.0 | 1480 | 186 | $0.75 / $3.8 | 1.0M | |
| 35 | Kimi K3 (batch)OPEN moonshotai | 70.2 | 43.6 | — | — | $2.3 / $11 | 1.0M | |
| 36 | 70.0 | 33.6 | 1482 | 203 | $1.5 / $9.0 | 1.0M | ||
| 37 | meta | 69.9 | 33.7 | 1479 | 119 | $1.3 / $4.3 | 1.0M | |
| 38 | anthropic | 69.9 | 38.2 | 1443 | 86 | $2.0 / $10 | 1M | |
| 39 | DeepSeek V4 Pro 0423OPEN deepseek | 69.4 | 36.0 | 1451 | 105 | $0.21 / $0.42 | 1.0M | |
| 40 | DeepSeek V4 Pro 0813OPEN deepseek | 69.4 | 36.0 | 1451 | 105 | $0.55 / $5.0 | 1.0M | |
| 41 | DeepSeek V4.1 FlashOPEN deepseek | 69.3 | 39.5 | 1463 | 213 | $0.00 / $2.4 | 1.0M | |
| 42 | GPT-6 Sol (batch) openai | 68.0 | 47.6 | — | — | $1.0 / $5.0 | 1.1M | |
| 43 | GLM 5.3 Flash (batch)OPEN z-ai | 67.9 | 41.8 | — | — | $0.06 / $0.20 | 1.0M | |
| 44 | Qwen3.8 27BOPEN qwen | 67.9 | 33.7 | 1441 | 45 | $0.42 / $2.5 | 1M | |
| 45 | MiMo-V2.6-FlashOPEN xiaomi | 67.4 | 37.9 | 1456 | 52 | $0.14 / $0.28 | 1.1M | |
| 46 | GPT-5.6 Terra (batch) openai | 66.8 | 42.1 | — | — | $1.0 / $6.0 | 1.1M | |
| 47 | openai | 66.8 | 42.1 | — | 111 | $2.0 / $12 | 1.1M | |
| 48 | Qwen3.7 Max qwen | 66.6 | 29.5 | 1476 | 198 | $1.5 / $4.4 | 1M | |
| 49 | Qwen3.8 2.4T A95BOPEN qwen | 66.4 | 39.9 | — | 38 | $2.0 / $6.0 | 1.0M | |
| 50 | DeepSeek V4 Flash 0731OPEN deepseek | 66.3 | 34.3 | 1432 | 215 | $0.01 / $1.3 | 1.0M | |
| 51 | deepseek | 66.3 | 34.3 | 1432 | 215 | $0.03 / $1.3 | 1.0M | |
| 52 | Grok 4.7 x-ai | 66.3 | 46.4 | — | 83 | $2.0 / $6.0 | 500k | |
| 53 | GPT-6 Sol openai | 66.3 | 47.6 | 1395 | 107 | $2.0 / $10 | 1.1M | |
| 54 | Claude Opus 4.8 (batch) anthropic | 65.5 | 41.8 | — | — | $2.5 / $13 | 1M | |
| 55 | anthropic | 65.3 | 26.4 | 1503 | 37 | $5.0 / $25 | 1M | |
| 56 | 65.1 | 29.7 | 1480 | 119 | $2.0 / $12 | 1.0M | ||
| 57 | Qwen3.5-Flash qwen | 65.0 | — | 1397 | — | $0.07 / $0.26 | 1M | |
| 58 | Gemini 3.8 Flash (batch) google | 64.8 | 40.9 | — | — | $0.38 / $1.9 | 1.0M | |
| 59 | Muse Spark 1.2 meta | 63.8 | 39.6 | — | 252 | $1.3 / $4.3 | 1.0M | |
| 60 | anthropic | 63.7 | 24.7 | 1458 | 43 | $3.0 / $15 | 1M | |
| 61 | MiMo-V2.5-ProOPEN xiaomi | 62.9 | 26.0 | 1465 | 44 | $0.43 / $0.87 | 1.1M | |
| 62 | Claude Sonnet 5 (batch) anthropic | 62.7 | 38.2 | — | — | $1.0 / $5.0 | 1M | |
| 63 | GLM 5.1OPEN z-ai | 62.3 | 26.1 | 1462 | 57 | $1.4 / $4.4 | 205k | |
| 64 | Kimi K2.6OPEN moonshotai | 62.1 | 27.0 | 1456 | 60 | $0.95 / $4.0 | 262k | |
| 65 | Gemini 3.7 Flash (batch) google | 61.6 | 39.1 | — | — | $0.38 / $1.9 | 1.0M | |
| 66 | GPT-5.6 Luna (batch) openai | 61.5 | 37.3 | — | — | $0.10 / $0.60 | 1.1M | |
| 67 | openai | 61.5 | 37.3 | — | 128 | $0.20 / $1.2 | 1.1M | |
| 68 | GPT-5.5 (batch) openai | 61.2 | 38.4 | — | — | $2.5 / $15 | 1.1M | |
| 69 | MiniMax M3OPEN minimax | 61.0 | 29.2 | 1432 | 97 | $0.30 / $1.2 | 1.0M | |
| 70 | Qwen3.6 Plus qwen | 60.7 | 27.0 | 1437 | 56 | $0.33 / $1.9 | 1M | |
| 71 | Gemini 3 Flash Preview google | 60.3 | 26.3 | 1466 | 191 | $0.50 / $3.0 | 1.0M | |
| 72 | Qwen3.7 Plus qwen | 60.0 | 25.2 | 1455 | 56 | $0.32 / $1.3 | 1M | |
| 73 | DeepSeek V4 Flash Vision ExpOPEN deepseek | 59.9 | 34.8 | — | 212 | $0.22 / $0.65 | 1.0M | |
| 74 | Hy3OPEN tencent | 59.8 | 25.3 | 1441 | 88 | $0.13 / $0.53 | 262k | |
| 75 | Hy3 previewOPEN tencent | 59.8 | 25.3 | 1441 | 88 | $0.18 / $0.60 | 262k | |
| 76 | GPT-5.1 openai | 59.4 | 24.7 | 1443 | 129 | $1.3 / $10 | 400k | |
| 77 | Qwen3.8 27B (free)OPEN qwen | 59.3 | 33.7 | — | — | Free / Free | 262k | |
| 78 | GPT-6 Luna openai | 59.0 | 38.1 | 1392 | 141 | $0.10 / $0.50 | 1.1M | |
| 79 | GLM 5OPEN z-ai | 58.9 | 27.9 | 1446 | 79 | $0.60 / $1.9 | 205k | |
| 80 | Kimi K2.5OPEN moonshotai | 58.7 | 23.5 | 1445 | — | $0.45 / $2.3 | 262k | |
| 81 | Ling 3.1 Flash inclusionai | 58.7 | 41.1 | — | 211 | Free / Free | 262k | |
| 82 | InklingOPEN thinkingmachines | 58.6 | 25.0 | 1442 | 172 | $0.95 / $4.0 | 524k | |
| 83 | openai | 56.7 | 30.4 | 1416 | 84 | $1.8 / $14 | 400k | |
| 84 | GPT-5.2 Chat openai | 56.7 | 30.4 | 1416 | 84 | $1.8 / $14 | 128k | |
| 85 | GLM 4.7OPEN z-ai | 56.5 | 22.2 | 1436 | 65 | $0.60 / $2.2 | 205k | |
| 86 | DeepSeek V4.1 Flash (batch)OPEN deepseek | 56.4 | 39.5 | — | — | $0.11 / $0.34 | 1.0M | |
| 87 | Nova 2 Lite amazon | 55.7 | — | 1362 | — | $0.30 / $2.5 | 1M | |
| 88 | Inkling SmallOPEN thinkingmachines | 55.4 | 25.7 | 1414 | 191 | $0.45 / $1.2 | 524k | |
| 89 | GLM 4.6OPEN z-ai | 55.4 | 18.5 | 1440 | 35 | $0.43 / $1.8 | 205k | |
| 90 | Gemini 3.5 Flash Lite google | 54.8 | 22.2 | 1435 | 324 | $0.30 / $2.5 | 1.0M | |
| 91 | DeepSeek V3.2OPEN deepseek | 54.6 | 21.5 | 1425 | — | $0.28 / $0.42 | 164k | |
| 92 | DeepSeek V3.2 ExpOPEN deepseek | 54.6 | 21.5 | 1425 | — | $0.27 / $0.41 | 164k | |
| 93 | GLM 5V Turbo z-ai | 54.6 | 23.5 | 1437 | — | $1.2 / $4.0 | 203k | |
| 94 | GPT-6 Luna (batch) openai | 54.4 | 38.1 | — | — | $0.05 / $0.25 | 1.1M | |
| 95 | Solar Pro 4 upstage | 54.3 | 28.2 | 1386 | 110 | $0.09 / $0.36 | 524k | |
| 96 | GPT-5.4 Mini openai | 54.0 | 24.1 | 1411 | 240 | $0.75 / $4.5 | 400k | |
| 97 | Gemini 3.6 Flash (batch) google | 53.9 | 34.0 | — | — | $0.38 / $1.9 | 1.0M | |
| 98 | Qwen3 Max Thinking qwen | 53.2 | 21.3 | 1439 | — | $0.78 / $3.9 | 262k | |
| 99 | Qwen3.5 397B A17BOPEN qwen | 52.6 | 18.4 | 1438 | 84 | $0.55 / $3.5 | 262k | |
| 100 | Gemini 3.5 Flash (batch) google | 52.0 | 32.6 | — | — | $0.75 / $4.5 | 1.0M | |
| 101 | MiniMax M2.7OPEN minimax | 51.7 | 22.8 | 1405 | 65 | $0.21 / $0.84 | 205k | |
| 102 | openai | 51.6 | 23.0 | 1406 | 115 | $1.3 / $10 | 400k | |
| 103 | Gemini 2.5 Pro google | 51.2 | 16.1 | 1458 | 128 | $1.3 / $10 | 1.0M | |
| 104 | Gemini 2.5 Pro Preview 06-05 google | 51.2 | 16.1 | 1458 | 128 | $1.3 / $10 | 1.0M | |
| 105 | MiniMax M1 minimax | 50.7 | — | 1343 | — | $0.55 / $2.2 | 1M | |
| 106 | Gemma 4 31BOPEN google | 50.6 | 14.7 | 1443 | 36 | $0.09 / $0.34 | 262k | |
| 107 | Qwen3.5-27BOPEN qwen | 50.4 | 22.9 | 1409 | 76 | $0.20 / $1.6 | 262k | |
| 108 | Claude Sonnet 4.6 (batch) anthropic | 50.4 | 30.1 | — | — | $1.5 / $7.5 | 1M | |
| 109 | x-ai | 50.0 | 24.9 | 1397 | 145 | $1.3 / $2.5 | 1M | |
| 110 | Gemma 4 26B A4B OPEN google | 49.3 | 16.7 | 1434 | — | $0.07 / $0.23 | 262k | |
| 111 | Qwen3 Max qwen | 49.2 | 15.6 | 1439 | 64 | $0.78 / $3.9 | 262k | |
| 112 | Mistral Medium 3.5 mistralai | 48.5 | 14.2 | 1421 | 164 | $1.5 / $7.5 | 262k | |
| 113 | Qwen3.5-122B-A10BOPEN qwen | 48.4 | 15.6 | 1417 | 129 | $0.26 / $2.1 | 262k | |
| 114 | Qwen3.5-35B-A3BOPEN qwen | 48.2 | 19.3 | 1394 | 143 | $0.15 / $1.0 | 262k | |
| 115 | DeepSeek V3.1 TerminusOPEN deepseek | 48.1 | 14.8 | 1420 | — | $0.27 / $1.0 | 164k | |
| 116 | GPT-5.4 Nano openai | 47.3 | 20.7 | 1372 | 170 | $0.20 / $1.3 | 400k | |
| 117 | MiniMax M2.1OPEN minimax | 46.8 | 20.9 | 1392 | 96 | $0.30 / $1.2 | 205k | |
| 118 | GPT-5.3-Codex openai | 46.4 | 32.5 | — | 102 | $1.8 / $14 | 400k | |
| 119 | GLM 4.5OPEN z-ai | 46.0 | 12.8 | 1430 | — | $0.60 / $2.2 | 131k | |
| 120 | R1 0528OPEN deepseek | 45.9 | 13.1 | 1428 | — | $0.50 / $2.1 | 164k | |
| 121 | Gemini 3.1 Flash Lite google | 45.6 | 15.6 | 1416 | 273 | $0.25 / $1.5 | 1.0M | |
| 122 | Gemini 3.1 Flash Lite Preview google | 45.6 | 15.6 | 1416 | 273 | $0.25 / $1.5 | 1.0M | |
| 123 | Step 3.5 FlashOPEN stepfun | 45.4 | 17.0 | 1403 | 114 | $0.10 / $0.30 | 262k | |
| 124 | DeepSeek V3.1OPEN deepseek | 45.3 | 13.7 | 1420 | — | $0.25 / $0.95 | 164k | |
| 125 | Qwen3 VL 235B A22B InstructOPEN qwen | 45.1 | 13.4 | 1420 | 58 | $0.21 / $1.9 | 262k | |
| 126 | Qwen3 VL 235B A22B ThinkingOPEN qwen | 45.1 | 13.4 | 1420 | 58 | $0.40 / $4.0 | 131k | |
| 127 | Kimi K2 0711OPEN moonshotai | 44.8 | 22.0 | 1371 | 159 | $0.57 / $2.3 | 131k | |
| 128 | Gemini 2.5 Flash google | 44.5 | 13.1 | 1417 | 215 | $0.30 / $2.5 | 1.0M | |
| 129 | MiniMax M2.5OPEN minimax | 43.8 | 22.8 | 1359 | 94 | $0.27 / $1.1 | 205k | |
| 130 | Ling 3.0 Flash VLOPEN inclusionai | 43.7 | 24.6 | — | — | $0.02 / $0.06 | 262k | |
| 131 | Qwen3 Next 80B A3B InstructOPEN qwen | 43.7 | 11.2 | 1417 | 194 | $0.10 / $1.1 | 262k | |
| 132 | Qwen3 Next 80B A3B ThinkingOPEN qwen | 43.7 | 11.2 | 1417 | 194 | $0.15 / $1.2 | 262k | |
| 133 | Qwen3 235B A22B Instruct 2507OPEN qwen | 43.3 | 12.7 | 1419 | 65 | $0.09 / $0.35 | 262k | |
| 134 | Gemini 3.1 Pro Preview (batch) google | 43.3 | 29.7 | — | — | $1.0 / $6.0 | 1.0M | |
| 135 | Kimi K2.7 CodeOPEN moonshotai | 42.8 | 25.8 | — | 97 | $0.67 / $3.4 | 262k | |
| 136 | Ling 3.0 Flash Fin inclusionai | 41.6 | 22.6 | — | 338 | $0.04 / $0.12 | 262k | |
| 137 | Qwen3 235B A22B Thinking 2507OPEN qwen | 41.6 | 9.5 | 1415 | 63 | $0.23 / $2.3 | 131k | |
| 138 | Claude Opus 4.5 anthropic | 41.6 | 29.1 | — | 47 | $5.0 / $25 | 200k | |
| 139 | Inkling Small (free)OPEN thinkingmachines | 41.4 | 25.7 | — | — | Free / Free | 1.0M | |
| 140 | o1 openai | 41.3 | 15.2 | 1353 | — | $15 / $60 | 200k | |
| 141 | Kimi K2 0905OPEN moonshotai | 41.2 | 15.3 | 1380 | 54 | $0.60 / $2.5 | 262k | |
| 142 | GPT-5.2-Codex openai | 40.7 | 28.5 | — | — | $1.8 / $14 | 400k | |
| 143 | Inkling (free)OPEN thinkingmachines | 40.4 | 25.0 | — | — | Free / Free | 1.0M | |
| 144 | GPT-5 Mini openai | 40.1 | 20.6 | 1373 | 136 | $0.25 / $2.0 | 400k | |
| 145 | GPT-5.4 Mini (batch) openai | 39.2 | 24.1 | — | — | $0.38 / $2.3 | 400k | |
| 146 | GLM 4.5 AirOPEN z-ai | 38.7 | 11.1 | 1384 | 80 | $0.13 / $0.85 | 131k | |
| 147 | MiniMax M2OPEN minimax | 38.3 | 18.6 | 1340 | 97 | $0.30 / $1.2 | 205k | |
| 148 | GLM 5 Turbo z-ai | 38.0 | 26.6 | — | — | $1.2 / $4.0 | 203k | |
| 149 | GLM 4.6VOPEN z-ai | 37.8 | 11.2 | 1377 | 89 | $0.30 / $0.90 | 131k | |
| 150 | R1OPEN deepseek | 37.6 | 13.1 | 1373 | — | $0.70 / $2.5 | 64k | |
| 151 | gpt-oss-120bOPEN openai | 37.4 | 11.6 | 1365 | 196 | $0.04 / $0.17 | 131k | |
| 152 | Nemotron 3 Ultra (free)OPEN nvidia | 37.2 | 22.9 | — | — | Free / Free | 1M | |
| 153 | Nemotron 3 UltraOPEN nvidia | 37.2 | 22.9 | — | — | $0.50 / $2.2 | 262k | |
| 154 | GLM 4.7 FlashOPEN z-ai | 37.1 | 14.9 | 1351 | 81 | $0.06 / $0.40 | 200k | |
| 155 | Qwen3.6 27BOPEN qwen | 36.8 | 21.4 | — | 56 | $0.32 / $2.7 | 262k | |
| 156 | x-ai | 36.7 | 25.7 | — | 113 | $1.3 / $2.5 | 2M | |
| 157 | Mercury 2 inception | 36.5 | 13.8 | 1355 | 722 | $0.25 / $0.75 | 128k | |
| 158 | Qwen3 30B A3B Instruct 2507OPEN qwen | 36.2 | 7.6 | 1384 | 109 | $0.05 / $0.19 | 262k | |
| 159 | GPT-5.4 Nano (batch) openai | 36.1 | 20.7 | — | — | $0.10 / $0.63 | 400k | |
| 160 | DeepSeek V3 0324OPEN deepseek | 36.1 | 9.7 | 1375 | — | $0.29 / $1.1 | 164k | |
| 161 | Qwen3 Coder 480B A35BOPEN qwen | 35.6 | 11.9 | 1356 | 50 | $0.30 / $1.0 | 262k | |
| 162 | Ling 3.0 FlashOPEN inclusionai | 35.3 | 20.1 | — | 347 | $0.02 / $0.06 | 262k | |
| 163 | Qwen3 235B A22BOPEN qwen | 35.3 | 9.5 | 1367 | 63 | $0.46 / $1.8 | 131k | |
| 164 | Grok 4.3 (batch) x-ai | 35.0 | 24.9 | — | — | $1.0 / $2.0 | 1M | |
| 165 | Claude Sonnet 4.5 (batch) anthropic | 34.9 | 20.7 | — | — | $1.5 / $7.5 | 1M | |
| 166 | Claude Sonnet 4.5 anthropic | 34.9 | 20.7 | — | — | $3.0 / $15 | 1M | |
| 167 | Gemini 3.5 Flash Lite (batch) google | 34.7 | 22.2 | — | — | $0.15 / $1.3 | 1.0M | |
| 168 | Solar Mini 4 upstage | 34.4 | 24.1 | — | — | $0.05 / $0.20 | 524k | |
| 169 | GPT-5.1-Codex-Max openai | 33.9 | 23.7 | — | — | $1.3 / $10 | 400k | |
| 170 | GPT-5.1-Codex openai | 33.9 | 23.7 | — | — | $1.3 / $10 | 400k | |
| 171 | o3 Mini openai | 33.5 | 12.5 | 1337 | 207 | $1.1 / $4.4 | 200k | |
| 172 | Step 3.7 FlashOPEN stepfun | 33.2 | 19.5 | — | 191 | $0.20 / $1.1 | 262k | |
| 173 | Trinity Large ThinkingOPEN arcee-ai | 32.6 | 10.8 | 1340 | 335 | $0.25 / $0.80 | 262k | |
| 174 | Llama 3.3 70B InstructOPEN meta-llama | 32.6 | — | 1274 | — | $0.22 / $0.50 | 131k | |
| 175 | GPT-5 Nano openai | 31.7 | 13.0 | 1320 | 158 | $0.05 / $0.40 | 400k | |
| 176 | o3 Mini High openai | 31.4 | 12.5 | 1337 | 207 | $1.1 / $4.4 | 200k | |
| 177 | o3 Pro openai | 31.3 | 21.9 | — | — | $20 / $80 | 200k | |
| 178 | LongCat 2.0OPEN meituan | 31.2 | 19.1 | — | — | $0.30 / $1.2 | 1.0M | |
| 179 | Gemma 3 27BOPEN google | 30.5 | 4.9 | 1358 | — | $0.08 / $0.45 | 131k | |
| 180 | Qwen3 32BOPEN qwen | 30.4 | 8.6 | 1340 | 107 | $0.08 / $0.28 | 131k | |
| 181 | DeepSeek V3OPEN deepseek | 30.4 | 8.5 | 1333 | — | $0.26 / $1.0 | 164k | |
| 182 | GLM 4.5VOPEN z-ai | 29.5 | 7.6 | 1333 | 43 | $0.60 / $1.8 | 66k | |
| 183 | Granite 4.2 8BOPEN ibm-granite | 29.4 | 11.1 | 1318 | 51 | $0.06 / $0.25 | 131k | |
| 184 | Qwen3.6 35B A3BOPEN qwen | 29.3 | 18.2 | — | 130 | $0.15 / $1.0 | 262k | |
| 185 | Llama 3.1 70B InstructOPEN meta-llama | 29.2 | — | 1261 | — | $0.40 / $0.40 | 131k | |
| 186 | GPT-5.1-Codex-Mini openai | 29.1 | 20.4 | — | — | $0.25 / $2.0 | 400k | |
| 187 | o3 openai | 28.9 | 20.2 | — | 138 | $2.0 / $8.0 | 200k | |
| 188 | Kimi K2 ThinkingOPEN moonshotai | 28.7 | 22.0 | — | 159 | $0.60 / $2.5 | 262k | |
| 189 | GPT-4o (2024-05-13) openai | 27.7 | 7.3 | 1300 | 130 | $5.0 / $15 | 128k | |
| 190 | Qwen3 30B A3BOPEN qwen | 27.3 | 7.6 | 1317 | 109 | $0.12 / $0.50 | 131k | |
| 191 | Claude Haiku 4.5 (batch) anthropic | 27.1 | 16.9 | — | — | $0.50 / $2.5 | 200k | |
| 192 | anthropic | 27.1 | 16.9 | — | — | $1.0 / $5.0 | 200k | |
| 193 | Qwen3 30B A3B Thinking 2507OPEN qwen | 26.5 | 7.6 | 1317 | 109 | $0.20 / $2.4 | 82k | |
| 194 | Gemma 3 12BOPEN google | 26.4 | 3.8 | 1334 | — | $0.05 / $0.15 | 131k | |
| 195 | Mistral Medium 3.5 (batch) mistralai | 25.7 | 14.2 | — | — | $0.75 / $3.8 | 262k | |
| 196 | gpt-oss-20bOPEN openai | 24.8 | 10.0 | 1288 | 257 | $0.02 / $0.09 | 131k | |
| 197 | Gemma 4 31B (free)OPEN google | 24.1 | 14.7 | — | — | Free / Free | 262k | |
| 198 | o4 Mini High openai | 23.9 | 16.7 | — | 155 | $1.1 / $4.4 | 200k | |
| 199 | o4 Mini openai | 23.9 | 16.7 | — | 155 | $1.1 / $4.4 | 200k | |
| 200 | GPT-4o-mini (2024-07-18) openai | 23.2 | 6.7 | 1286 | 90 | $0.15 / $0.60 | 128k | |
| 201 | GPT-4o (2024-08-06) openai | 22.9 | 7.7 | 1283 | 124 | $2.5 / $10 | 128k | |
| 202 | Gemma 2 27BOPEN google | 21.4 | — | 1231 | — | $0.65 / $0.65 | 8k | |
| 203 | Gemma 3 4BOPEN google | 21.3 | 4.8 | 1291 | — | $0.05 / $0.10 | 131k | |
| 204 | Gemini 2.5 Pro (batch) google | 21.3 | 16.1 | — | — | $0.63 / $5.0 | 1.0M | |
| 205 | Command R+ (08-2024) cohere | 20.8 | — | 1229 | — | $2.5 / $10 | 128k | |
| 206 | Nemotron 3 SuperOPEN nvidia | 20.2 | 12.8 | — | — | $0.08 / $0.45 | 262k | |
| 207 | Nemotron 3 Super (free)OPEN nvidia | 20.2 | 12.8 | — | — | Free / Free | 262k | |
| 208 | Mistral Large 2407 mistralai | 20.1 | 6.8 | 1266 | — | $2.0 / $6.0 | 131k | |
| 209 | GPT-5 Mini (batch) openai | 18.8 | 16.8 | — | — | $0.13 / $1.0 | 400k | |
| 210 | gpt-oss-120b (batch)OPEN openai | 18.1 | 11.6 | — | — | $0.03 / $0.14 | 131k | |
| 211 | GPT-4.1 openai | 18.1 | 12.7 | — | 163 | $2.0 / $8.0 | 1.0M | |
| 212 | Nemotron 3.5 LightningOPEN nvidia | 17.9 | 12.9 | — | 299 | $0.06 / $0.16 | 262k | |
| 213 | Nemotron 3.5 Lightning (free)OPEN nvidia | 17.9 | 12.9 | — | — | Free / Free | 1M | |
| 214 | o1-pro openai | 17.7 | 12.4 | — | — | $150 / $600 | 200k | |
| 215 | Mercury 2.5 inception | 17.6 | 12.3 | — | 621 | $0.04 / $0.15 | 260k | |
| 216 | North Mini Code (free)OPEN cohere | 17.6 | 9.9 | — | — | Free / Free | 256k | |
| 217 | Command A+ cohere | 17.4 | 13.1 | — | 156 | $0.30 / $1.5 | 192k | |
| 218 | GPT-4.1 Mini openai | 17.2 | 10.2 | — | 179 | $0.40 / $1.6 | 1.0M | |
| 219 | Qwen3 VL 32B InstructOPEN qwen | 17.0 | 11.9 | — | 91 | $0.10 / $0.42 | 131k | |
| 220 | Qwen3 Coder NextOPEN qwen | 16.9 | 9.2 | — | 100 | $0.12 / $0.80 | 262k | |
| 221 | Sonar Reasoning Pro perplexity | 16.9 | 11.8 | — | — | $2.0 / $8.0 | 128k | |
| 222 | Qwen3.5-9BOPEN qwen | 16.4 | 11.2 | — | 67 | $0.10 / $0.15 | 262k | |
| 223 | Devstral 2 2512OPEN mistralai | 15.9 | 9.0 | — | — | $0.40 / $2.0 | 262k | |
| 224 | Mistral Small 3OPEN mistralai | 15.8 | 6.7 | 1234 | — | $0.05 / $0.08 | 33k | |
| 225 | Mistral Small 4OPEN mistralai | 15.7 | 11.3 | — | 174 | $0.15 / $0.60 | 262k | |
| 226 | Mistral Small 4 (batch)OPEN mistralai | 15.7 | 11.3 | — | — | $0.07 / $0.30 | 262k | |
| 227 | GPT-4 Turbo openai | 14.6 | 7.0 | — | 33 | $10 / $30 | 128k | |
| 228 | Qwen3 Coder 30B A3B InstructOPEN qwen | 13.7 | 9.6 | — | 87 | $0.07 / $0.28 | 262k | |
| 229 | Qwen3 VL 30B A3B InstructOPEN qwen | 13.6 | 9.5 | — | 113 | $0.15 / $0.60 | 262k | |
| 230 | Qwen3 VL 30B A3B ThinkingOPEN qwen | 13.6 | 9.5 | — | 113 | $0.20 / $2.4 | 262k | |
| 231 | Command AOPEN cohere | 13.1 | 7.0 | — | 64 | $2.5 / $10 | 256k | |
| 232 | Phi 4OPEN microsoft | 13.0 | 5.9 | 1217 | 38 | $0.07 / $0.14 | 16k | |
| 233 | Mistral Medium 3.1 (batch) mistralai | 12.9 | 9.2 | — | — | $0.20 / $1.0 | 131k | |
| 234 | Mistral Medium 3.1 mistralai | 12.9 | 9.2 | — | — | $0.40 / $2.0 | 131k | |
| 235 | Mistral Medium 3 mistralai | 12.9 | 9.0 | — | — | $0.40 / $2.0 | 131k | |
| 236 | gpt-oss-20b (batch)OPEN openai | 12.6 | 9.0 | — | — | $0.02 / $0.11 | 131k | |
| 237 | Mistral Large 3 2512 (batch) mistralai | 12.6 | 9.3 | — | — | $0.25 / $0.75 | 262k | |
| 238 | Sonar perplexity | 12.4 | 8.7 | — | — | $1.0 / $1.0 | 127k | |
| 239 | Gemini 2.5 Flash Lite google | 12.1 | 8.5 | — | 354 | $0.10 / $0.40 | 1.0M | |
| 240 | GPT-4o (2024-11-20) openai | 12.0 | 8.4 | — | 140 | $2.5 / $10 | 128k | |
| 241 | GPT-4o openai | 12.0 | 8.4 | — | 140 | $2.5 / $10 | 128k | |
| 242 | Llama 4 MaverickOPEN meta-llama | 11.9 | 10.0 | — | 69 | $0.19 / $0.65 | 1.0M | |
| 243 | Qwen3 VL 8B InstructOPEN qwen | 11.7 | 8.2 | — | 115 | $0.12 / $0.46 | 262k | |
| 244 | Qwen3 VL 8B ThinkingOPEN qwen | 11.7 | 8.2 | — | 115 | $0.18 / $2.1 | 131k | |
| 245 | GPT-4.1 Nano openai | 11.5 | 7.8 | — | 187 | $0.10 / $0.40 | 1.0M | |
| 246 | GPT-4 openai | 11.2 | 6.7 | — | — | $30 / $60 | 8k | |
| 247 | Sonar Pro perplexity | 10.9 | 7.6 | — | — | $3.0 / $15 | 200k | |
| 248 | Nemotron 3 Nano 30B A3BOPEN nvidia | 10.7 | 8.9 | — | — | $0.05 / $0.20 | 262k | |
| 249 | Solar Pro 3 upstage | 10.6 | 7.8 | — | 136 | $0.15 / $0.60 | 131k | |
| 250 | GPT-4o-mini openai | 10.6 | 6.7 | — | 90 | $0.15 / $0.60 | 128k |
The raw measurements come from third parties, credited below. What we add is the cross-source aggregation and the scoring: five axes on absolute 0-100 scales, weighted into one number. Absolute scales matter — a model scoring 78 today means the same as a model scoring 78 six months ago, so the history stays comparable as new models arrive.
Average of every published capability index we can cross-check — general intelligence, coding and agentic benchmarks.
Real market price per million tokens weighted against measured capability. A cheap weak model and an expensive strong one can score the same.
Context window, maximum output, modalities, reasoning support, prompt caching and how recent the knowledge cutoff is.
Measured output throughput (tokens per second) and time to first token. A reasoning model thinks before it answers, so its first token can take minutes — that is the trade it makes for capability, and it scores low here by design.
How many providers serve the model, and whether the weights are open — availability and independence from a single vendor.
Every collection is archived, so price cuts and ranking moves stay traceable. The sources publish today's numbers; this is the trail behind them.
The default ranking is capability, so the most expensive model can sit above a free one. Switch the criterion at the top of the page and the order changes completely — that is the point. Four questions come up constantly, and here are the answers.
Because the default ranking measures what a model can do, not what it costs. Price is a separate axis, and it has its own ranking one click away. Ranking on the composite score instead buried the frontier models behind cheap ones and turned the whole index into a price chart, which is not what a ranked list of models is for. Today that puts Anthropic: Claude Opus 5.5 first with 88.7 on capability at $20.00 per million output tokens.
Both count, equally. Benchmarks test verifiable problems; the arena records which answer people actually preferred. When a model carries both, each contributes half its intelligence score. The widest gap in the index right now is Google: Gemini 2.5 Pro: 23 from benchmarks against 81 from human votes — a 58-point disagreement that a single-source ranking would hide entirely.
The score itself, split into the five things we weigh. Each segment is that axis's contribution, so the full bar length is the composite score. Two models can share a total and have opposite shapes — one fast and cheap, the other slow and brilliant.
Because that measurement does not exist yet, and we would rather show a dash than invent a number. Of the 250 models listed, 241 carry benchmark indices and 137 carry human ratings. A missing value never counts as zero — it simply leaves that axis out of the model's score, and the “Double-validated” filter isolates the 128 models measured both ways.
A ranking of 250 large language models built by cross-referencing three independent sources: live market pricing from OpenRouter, benchmark indices from Artificial Analysis, and human preference votes from LMArena. We aggregate those measurements and compute our own score across five weighted axes. The raw measurements belong to their sources; the scoring and the ranking are ours.
By measured capability, not by price. A model that costs ten times more but answers harder questions ranks higher by default. Three alternative orderings are one click away — price/performance, speed, and the weighted composite score — because "best" means something different depending on whether you are paying the bill or waiting for the response.
Two things, weighted equally when both exist. Machine benchmarks contribute the general intelligence index (50%), the coding index (25%) and the agentic index (25%), each normalised against its own ceiling. Human preference contributes the LMArena Elo rating, mapped onto the same 0-100 scale. A model measured by only one of the two is scored on that one alone.
A model carrying both a machine benchmark score and a human preference rating. 128 of the 250 models in the index qualify. It matters because the two methods disagree more often than you would expect: benchmarks measure what is testable, the arena measures what people actually prefer after using the model.
They measure different things. Benchmarks test verifiable problems with known answers — maths, code that compiles, factual recall. Human votes reward tone, instruction-following, formatting and refusal behaviour, which no benchmark captures. A model can top the benchmarks and lose blind head-to-head votes, and the reverse happens just as often.
Because nearly every current model scored in the 90s on it, so it separated nothing. Worse, averaging it in rewarded models measured on maths and almost nothing else — one model with no coding or agentic score reached a higher intelligence score than models measured on everything. Artificial Analysis has since stopped publishing that index on the endpoint we use, which changes nothing for us: it was already out of the score.
Measured capability divided by the blended price per million tokens, weighted one part input to three parts output, which reflects typical real-world usage. The result is placed on a logarithmic scale because prices span four orders of magnitude. Free or unpriced models receive a neutral 50 rather than a perfect score, since they are not comparable.
Yes. They come from OpenRouter, which publishes live per-token rates charged by the providers actually serving each model, in US dollars per million tokens. Input and output are shown separately because the gap between them is often large and matters more than the headline figure.
Every 48 hours. Each collection re-fetches all three sources, recalculates every score, and writes a historical snapshot so price changes and ranking movements stay traceable. Models that disappear from the sources are marked inactive rather than deleted, so their history survives.
A model whose weights can be downloaded and self-hosted, as opposed to one reachable only through a vendor API. 116 of the 250 models qualify. Open weights add points on the ecosystem axis because they remove vendor lock-in, but a closed model served by many providers is not penalised to zero.
So the numbers stay comparable over time. If scores were normalised against the current field, every new model release would silently move every existing score, and a snapshot from six months ago would mean nothing. On an absolute scale, a model scoring 78 today means the same as a model scoring 78 last year.
28 of the 250 models link to a full review on this site, written after actually using the model. Those links appear directly on each model card and in the comparison table. The rest are scored from published measurements only, which is stated rather than hidden.
All raw metrics belong to their respective sources and are reproduced with attribution. The composite scores, weighting and rankings on this page are ThePlanetTools.ai's own computation and should be attributed to us, not to the data providers.
We're developers and SaaS builders who use these tools daily in production. Every review comes from hands-on experience building real products — DealPropFirm, ThePlanetIndicator, PropFirmsCodes, and many more. We don't just review tools — we build and ship with them every day.
Written and tested by developers who build with these tools daily.