Рейтинг моделей для математики
Таблица по открытым замерам оценок людей: оценка, интервал и число голосов; переключайте категорию и дату снимка.
Динамика рейтинга
Загружаем историю замеров…
Как считается
Оценки и интервалы приведены как в открытых замерах оценок людей, без пересчёта; доля использования — по публичным данным API-агрегаторов (топ-K моделей и строка «прочие»); характеристики и цены — из каталогов и провайдеров с датой наблюдения. Интервал у оценки не равен диапазону мест: соседние модели с пересекающимися интервалами статистически неразличимы.
Полная методика: источники, частота обновления, что означает число
Смысл числа
- Метрика
- рейтинг Брэдли — Терри (Elo-подобная шкала) с интервалом из источника
- Единица выборки
- голос в парном сравнении
- Окно
- снимок датасета от 13 сентября 2026
- Как читать
- интервал не равен диапазону мест: модели с пересекающимися интервалами статистически неразличимы
Таблица борда
Снимок от 13 сентября 2026, категория «математика», 300 строк. Оценка и интервал — из источника; интервал не равен диапазону мест.
| Место | Модель | Разработчик | Оценка | Интервал | Голоса | Лицензия |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 в режиме max | Anthropic | 1 537 | 1 518–1 557 | 904 | Proprietary |
| 2 | Claude Opus 5 в режиме high | Anthropic | 1 534 | 1 520–1 549 | 1 866 | Proprietary |
| 3 | Gemini 3.7 Flash в режиме high | 1 525 | 1 492–1 558 | 310 | Proprietary | |
| 4 | Claude Fable 5.1 в режиме max | Anthropic | 1 522 | 1 485–1 558 | 243 | Proprietary |
| 5 | Claude Fable 5 | Anthropic | 1 522 | 1 506–1 537 | 1 467 | Proprietary |
| 6 | Claude Opus 4.6 в режиме high | Anthropic | 1 516 | 1 506–1 526 | 3 695 | Proprietary |
| 7 | Claude Opus 4.6 | Anthropic | 1 512 | 1 502–1 522 | 4 110 | Proprietary |
| 8 | GLM 5.3 Flash | Z.ai | 1 511 | 1 483–1 539 | 441 | MIT |
| 9 | Gemini 3.6 Flash в режиме high | 1 506 | 1 489–1 523 | 1 242 | Proprietary | |
| 10 | Gemini 3.5 Flash в режиме high | 1 505 | 1 491–1 519 | 1 867 | Proprietary | |
| 11 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 500 | 1 479–1 521 | 769 | Proprietary |
| 12 | GLM 5.3 в режиме max | Z.ai | 1 498 | 1 471–1 525 | 430 | MIT |
| 13 | Claude Opus 4.7 в режиме high | Anthropic | 1 498 | 1 487–1 509 | 3 030 | Proprietary |
| 14 | Kimi K3 в режиме max | Moonshot AI | 1 494 | 1 474–1 514 | 884 | Kimi K3 license |
| 15 | Qwen3.7 Max | Alibaba Qwen | 1 493 | 1 454–1 533 | 218 | Proprietary |
| 16 | GPT-5.4 в режиме high | OpenAI | 1 490 | 1 479–1 501 | 3 179 | Proprietary |
| 17 | Claude Opus 4.7 | Anthropic | 1 489 | 1 477–1 500 | 3 144 | Proprietary |
| 18 | Muse Spark 1.3 в режиме max | Meta | 1 489 | 1 450–1 527 | 207 | Proprietary |
| 19 | GPT-5.5 · 5.5 | OpenAI | 1 486 | 1 476–1 497 | 3 301 | Proprietary |
| 20 | Claude Opus 4.8 в режиме high | Anthropic | 1 486 | 1 474–1 498 | 2 444 | Proprietary |
| 21 | Gemini 3.1 Pro Preview | 1 486 | 1 477–1 495 | 5 569 | Proprietary | |
| 22 | Gemini 3.5 Flash в режиме medium | 1 482 | 1 467–1 497 | 1 667 | Proprietary | |
| 23 | ERNIE 5.1 | Baidu | 1 479 | 1 466–1 493 | 1 886 | Proprietary |
| 24 | Inkling | Thinking Machines Lab | 1 477 | 1 460–1 495 | 1 120 | Apache 2.0 |
| 25 | MiMo-V2.5-Pro | Xiaomi | 1 477 | 1 466–1 488 | 2 879 | MIT |
| 26 | GPT-5.5 в режиме high | OpenAI | 1 476 | 1 465–1 486 | 3 261 | Proprietary |
| 27 | Gemini 3 Pro | 1 476 | 1 464–1 487 | 2 615 | Proprietary | |
| 28 | Qwen3.5 Max | Alibaba Qwen | 1 475 | 1 459–1 491 | 1 346 | Proprietary |
| 29 | GLM 5.2 в режиме max | Z.ai | 1 475 | 1 460–1 490 | 1 662 | MIT |
| 30 | Muse Spark 1.1 | Meta | 1 474 | 1 457–1 491 | 1 164 | Proprietary |
| 31 | Gemini 3 Flash Preview | 1 474 | 1 460–1 487 | 1 957 | Proprietary | |
| 32 | GLM 5.1 | Z.ai | 1 473 | 1 461–1 486 | 2 373 | MIT |
| 33 | Kimi K2.6 | Moonshot AI | 1 473 | 1 460–1 487 | 1 931 | Modified MIT |
| 34 | Hy3 | Tencent | 1 473 | 1 444–1 501 | 413 | Apache 2.0 |
| 35 | Qwen3.8 27B | Alibaba Qwen | 1 470 | 1 443–1 496 | 468 | Apache 2.0 |
| 36 | Kimi K2.5 · thinking | Moonshot AI | 1 469 | 1 460–1 479 | 3 916 | Modified MIT |
| 37 | Claude Sonnet 5 в режиме high | Anthropic | 1 469 | 1 454–1 485 | 1 519 | Proprietary |
| 38 | Claude Opus 4.8 | Anthropic | 1 468 | 1 456–1 481 | 2 491 | Proprietary |
| 39 | Gemma 4 31B | 1 468 | 1 441–1 496 | 403 | Apache 2.0 | |
| 40 | Gemma 4 26B A4B | 1 468 | 1 439–1 496 | 369 | Apache 2.0 | |
| 41 | Qwen3.7 Plus | Alibaba Qwen | 1 467 | 1 452–1 481 | 1 767 | Proprietary |
| 42 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 466 | 1 449–1 483 | 1 220 | Proprietary |
| 43 | Qwen3.6 Max Preview | Alibaba Qwen | 1 463 | 1 433–1 493 | 360 | Proprietary |
| 44 | Grok 4.5 | xAI | 1 462 | 1 445–1 479 | 1 232 | Proprietary |
| 45 | Claude Sonnet 4.6 | Anthropic | 1 462 | 1 451–1 472 | 3 551 | Proprietary |
| 46 | Claude Opus 4.5 в режиме high | Anthropic | 1 460 | 1 448–1 473 | 2 214 | Proprietary |
| 47 | Claude Opus 4.5 | Anthropic | 1 460 | 1 451–1 469 | 4 274 | Proprietary |
| 48 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 460 | 1 443–1 477 | 1 186 | Proprietary |
| 49 | GPT-5.4 | OpenAI | 1 458 | 1 447–1 468 | 3 345 | Proprietary |
| 50 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 457 | 1 445–1 470 | 2 498 | MIT |
| 51 | Grok 4.20 · beta-0309-reasoning | xAI | 1 457 | 1 447–1 468 | 3 315 | Proprietary |
| 52 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 457 | 1 441–1 474 | 1 273 | Proprietary |
| 53 | Nemotron 3 Ultra | NVIDIA | 1 456 | 1 432–1 481 | 547 | OpenMDW-1.1 |
| 54 | Qwen3.5 397B A17B | Alibaba Qwen | 1 451 | 1 442–1 461 | 4 244 | Apache 2.0 |
| 55 | Gemini 2.5 Pro | 1 451 | 1 444–1 458 | 7 555 | Proprietary | |
| 56 | Qwen3.6 Plus | Alibaba Qwen | 1 451 | 1 439–1 463 | 2 523 | Proprietary |
| 57 | Muse Spark | Meta | 1 450 | 1 430–1 470 | 857 | Proprietary |
| 58 | Inkling Small | Thinking Machines Lab | 1 450 | 1 428–1 472 | 722 | Apache 2.0 |
| 59 | Qwen3 Max · preview | Alibaba Qwen | 1 448 | 1 433–1 463 | 1 484 | Proprietary |
| 60 | MiMo-V2 Pro | Xiaomi | 1 448 | 1 434–1 463 | 1 608 | Proprietary |
| 61 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 448 | 1 439–1 456 | 4 834 | Proprietary |
| 62 | Gemini 3 Flash Preview в режиме minimal | 1 448 | 1 439–1 457 | 4 660 | Proprietary | |
| 63 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 446 | 1 415–1 477 | 340 | MIT |
| 64 | GLM 5V Turbo | Z.ai | 1 445 | 1 419–1 471 | 442 | Proprietary |
| 65 | GPT-5.1 в режиме high | OpenAI | 1 444 | 1 432–1 456 | 2 453 | Proprietary |
| 66 | DeepSeek V4 Pro 0423 | DeepSeek | 1 443 | 1 431–1 454 | 2 868 | MIT |
| 67 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 441 | 1 425–1 458 | 1 192 | Apache 2.0 |
| 68 | GPT-5.2 в режиме high | OpenAI | 1 441 | 1 430–1 452 | 2 950 | Proprietary |
| 69 | LongCat-Flash-Chat · chat | Meituan | 1 441 | 1 419–1 463 | 686 | MIT |
| 70 | Kimi K2.5 · instant | Moonshot AI | 1 440 | 1 415–1 465 | 507 | Modified MIT |
| 71 | Seed 2.0 Pro | ByteDance Seed | 1 440 | 1 430–1 449 | 4 071 | Proprietary |
| 72 | Grok 4.20 Multi-Agent | xAI | 1 439 | 1 428–1 450 | 3 222 | Proprietary |
| 73 | GLM 5 | Z.ai | 1 439 | 1 424–1 453 | 1 602 | MIT |
| 74 | ERNIE 5.0 · 0110 | Baidu | 1 438 | 1 426–1 451 | 2 118 | Proprietary |
| 75 | MiMo-V2.5 | Xiaomi | 1 438 | 1 425–1 450 | 2 327 | MIT |
| 76 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 437 | 1 398–1 477 | 207 | Proprietary |
| 77 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 437 | 1 426–1 448 | 2 957 | MIT |
| 78 | Grok 4.20 · beta1 | xAI | 1 436 | 1 421–1 450 | 1 593 | Proprietary |
| 79 | Gemini 3.5 Flash Lite | 1 436 | 1 418–1 453 | 1 140 | Proprietary | |
| 80 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 435 | 1 411–1 459 | 565 | Proprietary |
| 81 | GPT-5.2 Chat | OpenAI | 1 433 | 1 420–1 447 | 2 069 | Proprietary |
| 82 | Qwen3.5-27B | Alibaba Qwen | 1 433 | 1 418–1 447 | 1 640 | Apache 2.0 |
| 83 | Mistral Medium 3.5 | Mistral AI | 1 433 | 1 408–1 458 | 539 | Modified MIT |
| 84 | MiniMax M3 | MiniMax | 1 433 | 1 420–1 446 | 2 301 | MiniMax Community License |
| 85 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 432 | 1 418–1 447 | 1 748 | Proprietary |
| 86 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 432 | 1 418–1 447 | 1 542 | Proprietary |
| 87 | MiMo-V2 Omni | Xiaomi | 1 432 | 1 412–1 452 | 911 | Proprietary |
| 88 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 431 | 1 423–1 439 | 5 847 | Apache 2.0 |
| 89 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 431 | 1 420–1 441 | 2 993 | Proprietary |
| 90 | GLM 4.6 | Z.ai | 1 430 | 1 418–1 443 | 2 058 | MIT |
| 91 | Gemini 3.1 Flash Lite Preview | 1 430 | 1 419–1 440 | 3 402 | Proprietary | |
| 92 | Kimi K2 Thinking | Moonshot AI | 1 429 | 1 420–1 439 | 3 710 | Modified MIT |
| 93 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 429 | 1 415–1 443 | 1 764 | Apache 2.0 |
| 94 | GLM 4.5 | Z.ai | 1 428 | 1 412–1 444 | 1 408 | MIT |
| 95 | DeepSeek V4 Flash 0423 | DeepSeek | 1 427 | 1 415–1 439 | 2 514 | MIT |
| 96 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 427 | 1 407–1 447 | 803 | Proprietary |
| 97 | o3 | OpenAI | 1 426 | 1 416–1 436 | 3 679 | Proprietary |
| 98 | Muse Glimmer 30B | Meta | 1 425 | 1 384–1 466 | 206 | Apache-2.0 |
| 99 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 424 | 1 412–1 437 | 2 451 | MIT |
| 100 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 424 | 1 401–1 447 | 693 | Apache 2.0 |
| 101 | GLM 4.7 | Z.ai | 1 424 | 1 403–1 446 | 673 | MIT |
| 102 | Grok 4 | xAI | 1 424 | 1 411–1 436 | 2 221 | Proprietary |
| 103 | Grok 4.1 · 4.1-thinking | xAI | 1 423 | 1 414–1 433 | 3 756 | Proprietary |
| 104 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 423 | 1 396–1 450 | 470 | MIT |
| 105 | GPT-5.5 · 5.5-instant | OpenAI | 1 423 | 1 407–1 439 | 1 457 | Proprietary |
| 106 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 422 | 1 385–1 459 | 232 | Proprietary |
| 107 | Claude Sonnet 4.5 | Anthropic | 1 421 | 1 413–1 430 | 4 871 | Proprietary |
| 108 | Claude Opus 4.1 · 20250805 | Anthropic | 1 421 | 1 412–1 430 | 4 655 | Proprietary |
| 109 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 421 | 1 400–1 441 | 771 | MIT |
| 110 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 420 | 1 402–1 438 | 982 | MIT |
| 111 | MiniMax M2.7 | MiniMax | 1 420 | 1 409–1 430 | 3 454 | Modified MIT |
| 112 | Hy3 preview | Tencent | 1 420 | 1 392–1 447 | 405 | tencent-hunyuan-community |
| 113 | GPT-5.4 Mini в режиме high | OpenAI | 1 419 | 1 408–1 430 | 3 161 | Proprietary |
| 114 | Grok 4 Fast · chat | xAI | 1 419 | 1 390–1 448 | 383 | Proprietary |
| 115 | GPT-5.2 | OpenAI | 1 419 | 1 409–1 428 | 4 351 | Proprietary |
| 116 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 417 | 1 405–1 429 | 2 463 | MIT |
| 117 | Grok 4.1 · 4.1 | xAI | 1 417 | 1 407–1 426 | 4 165 | Proprietary |
| 118 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 416 | 1 394–1 438 | 658 | MIT |
| 119 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 415 | 1 402–1 429 | 1 927 | Proprietary | |
| 120 | Mistral Large 3 2512 | Mistral AI | 1 415 | 1 405–1 425 | 3 894 | Apache 2.0 |
| 121 | Grok 4.6 в режиме high | xAI | 1 415 | 1 393–1 437 | 719 | Proprietary |
| 122 | Hunyuan T1 | Tencent | 1 413 | 1 375–1 452 | 233 | Proprietary |
| 123 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 413 | 1 389–1 438 | 479 | Apache 2.0 |
| 124 | GPT-4.5 Preview | OpenAI | 1 412 | 1 397–1 427 | 1 393 | Proprietary |
| 125 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 412 | 1 383–1 440 | 416 | Apache 2.0 |
| 126 | Gemini 2.5 Flash · flash | 1 411 | 1 404–1 418 | 7 772 | Proprietary | |
| 127 | Mistral Medium 3.1 | Mistral AI | 1 410 | 1 403–1 418 | 5 749 | Proprietary |
| 128 | GPT-5.1 | OpenAI | 1 409 | 1 398–1 420 | 2 812 | Proprietary |
| 129 | ERNIE 5.0 · preview-1203 | Baidu | 1 409 | 1 385–1 432 | 604 | Proprietary |
| 130 | Qwen3.5-Flash | Alibaba Qwen | 1 408 | 1 397–1 419 | 3 191 | Proprietary |
| 131 | GPT-5 | OpenAI | 1 408 | 1 394–1 422 | 1 761 | Proprietary |
| 132 | ERNIE 5.0 · preview-1022 | Baidu | 1 407 | 1 373–1 441 | 265 | Proprietary |
| 133 | ChatGPT-4o (latest) | OpenAI | 1 406 | 1 399–1 414 | 5 650 | Proprietary |
| 134 | Step 3.5 Flash | StepFun | 1 406 | 1 396–1 417 | 3 224 | Apache 2.0 |
| 135 | GPT-5.4 Nano в режиме high | OpenAI | 1 405 | 1 394–1 416 | 3 072 | Proprietary |
| 136 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 405 | 1 391–1 419 | 1 751 | Apache 2.0 |
| 137 | Grok 4.1 Fast | xAI | 1 404 | 1 394–1 415 | 3 434 | Proprietary |
| 138 | Grok 4 Fast · reasoning | xAI | 1 404 | 1 386–1 422 | 1 053 | Proprietary |
| 139 | DeepSeek V3.1 Terminus | DeepSeek | 1 402 | 1 363–1 440 | 216 | MIT |
| 140 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 401 | 1 368–1 435 | 258 | Proprietary |
| 141 | R1 0528 | DeepSeek | 1 400 | 1 380–1 420 | 854 | MIT |
| 142 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 399 | 1 379–1 419 | 816 | Apache 2.0 |
| 143 | Qwen3 32B | Alibaba Qwen | 1 398 | 1 368–1 428 | 316 | Apache 2.0 |
| 144 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 398 | 1 386–1 410 | 2 363 | Apache 2.0 |
| 145 | GPT-5 в режиме high | OpenAI | 1 397 | 1 384–1 411 | 1 857 | Proprietary |
| 146 | GLM 4.5 Air | Z.ai | 1 397 | 1 382–1 412 | 1 511 | MIT |
| 147 | Kimi K2 0905 | Moonshot AI | 1 396 | 1 375–1 417 | 753 | Modified MIT |
| 148 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 396 | 1 385–1 407 | 2 797 | MIT |
| 149 | o3 Mini High | OpenAI | 1 396 | 1 382–1 409 | 1 909 | Proprietary |
| 150 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 394 | 1 381–1 408 | 1 597 | Apache 2.0 |
| 151 | MiniMax M2.1 | MiniMax | 1 394 | 1 376–1 413 | 963 | MIT |
| 152 | Claude Haiku 4.5 | Anthropic | 1 394 | 1 387–1 402 | 7 147 | Proprietary |
| 153 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 394 | 1 379–1 409 | 1 395 | Apache 2.0 |
| 154 | R1 | DeepSeek | 1 392 | 1 378–1 406 | 1 606 | MIT |
| 155 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 392 | 1 379–1 404 | 2 203 | Proprietary |
| 156 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 390 | 1 351–1 429 | 194 | Nvidia Open |
| 157 | Grok 3 | xAI | 1 390 | 1 379–1 401 | 2 665 | Proprietary |
| 158 | o4 Mini | OpenAI | 1 389 | 1 378–1 399 | 2 898 | Proprietary |
| 159 | o1 · 2024-12-17 | OpenAI | 1 388 | 1 377–1 399 | 2 986 | Proprietary |
| 160 | Grok 4.3 | xAI | 1 388 | 1 377–1 399 | 3 178 | Proprietary |
| 161 | gpt-oss-120b | OpenAI | 1 387 | 1 373–1 401 | 1 763 | Apache 2.0 |
| 162 | GPT-5.3 Chat | OpenAI | 1 385 | 1 372–1 399 | 2 028 | Proprietary |
| 163 | Grok 3 Mini в режиме high | xAI | 1 385 | 1 366–1 403 | 954 | Proprietary |
| 164 | Nemotron 3 Super | NVIDIA | 1 383 | 1 358–1 408 | 515 | NVIDIA Open Model |
| 165 | INTELLECT-3 | Prime Intellect | 1 382 | 1 351–1 413 | 332 | MIT |
| 166 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 382 | 1 359–1 404 | 610 | MIT |
| 167 | MiniMax M2.5 | MiniMax | 1 381 | 1 368–1 393 | 2 420 | Modified MIT |
| 168 | GPT-5 Mini в режиме high | OpenAI | 1 376 | 1 361–1 392 | 1 438 | Proprietary |
| 169 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 375 | 1 362–1 388 | 1 995 | Proprietary |
| 170 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 374 | 1 363–1 385 | 2 815 | Proprietary | |
| 171 | Claude Opus 4 · 20250514 | Anthropic | 1 374 | 1 363–1 385 | 2 722 | Proprietary |
| 172 | o3 Mini | OpenAI | 1 373 | 1 365–1 382 | 4 692 | Proprietary |
| 173 | DeepSeek V3 0324 | DeepSeek | 1 373 | 1 363–1 383 | 3 150 | MIT |
| 174 | o1 · preview | OpenAI | 1 373 | 1 363–1 383 | 4 569 | Proprietary |
| 175 | Grok 3 Mini | xAI | 1 373 | 1 358–1 387 | 1 490 | Proprietary |
| 176 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 370 | 1 351–1 388 | 950 | NVIDIA Open Model |
| 177 | Qwen2.5 Max | Alibaba Qwen | 1 369 | 1 359–1 379 | 3 298 | Proprietary |
| 178 | GPT-4.1 | OpenAI | 1 369 | 1 358–1 379 | 3 194 | Proprietary |
| 179 | Trinity Large Thinking | Arcee AI | 1 367 | 1 352–1 382 | 1 611 | Apache 2.0 |
| 180 | Ling-flash-2.0 | inclusionAI | 1 367 | 1 340–1 394 | 446 | MIT |
| 181 | Kimi K2 0711 | Moonshot AI | 1 366 | 1 352–1 381 | 1 679 | Modified MIT |
| 182 | Step 3 | StepFun | 1 365 | 1 334–1 397 | 348 | Apache 2.0 |
| 183 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 363 | 1 348–1 378 | 1 607 | Apache 2.0 |
| 184 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 362 | 1 350–1 375 | 2 057 | Proprietary | |
| 185 | Nova 2 Lite | Amazon | 1 361 | 1 340–1 381 | 820 | Proprietary |
| 186 | MiniMax M1 | MiniMax | 1 360 | 1 347–1 374 | 1 761 | Apache 2.0 |
| 187 | Claude Sonnet 4 · 20250514 | Anthropic | 1 360 | 1 348–1 371 | 2 428 | Proprietary |
| 188 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 1 359 | 1 322–1 396 | 209 | Nvidia Open Model |
| 189 | GLM 4.7 Flash | Z.ai | 1 359 | 1 338–1 381 | 704 | MIT |
| 190 | Hunyuan TurboS · 20250416 | Tencent | 1 359 | 1 340–1 378 | 845 | Proprietary |
| 191 | QwQ 32B · 32b | Alibaba Qwen | 1 358 | 1 345–1 372 | 1 704 | Apache 2.0 |
| 192 | o1-mini | OpenAI | 1 358 | 1 350–1 366 | 7 499 | Proprietary |
| 193 | Nemotron 3.5 Lightning | NVIDIA | 1 357 | 1 322–1 391 | 336 | OpenMDW-1.1 |
| 194 | GLM 4.5V | Z.ai | 1 355 | 1 321–1 389 | 271 | MIT |
| 195 | Qwen3 30B A3B | Alibaba Qwen | 1 354 | 1 341–1 368 | 1 692 | Apache 2.0 |
| 196 | MiniMax M2 | MiniMax | 1 353 | 1 320–1 386 | 318 | Apache 2.0 |
| 197 | Mistral Medium 3 | Mistral AI | 1 351 | 1 339–1 364 | 2 208 | Proprietary |
| 198 | Gemini 2.0 Flash | 1 351 | 1 342–1 360 | 4 045 | Proprietary | |
| 199 | Ring-flash-2.0 | inclusionAI | 1 344 | 1 317–1 371 | 445 | MIT |
| 200 | GPT-4.1 Mini | OpenAI | 1 342 | 1 331–1 354 | 2 666 | Proprietary |
| 201 | Mistral Small 3.2 24B | Mistral AI | 1 341 | 1 323–1 358 | 1 039 | Apache 2.0 |
| 202 | Trinity Large | Arcee AI | 1 336 | 1 322–1 351 | 1 871 | Apache 2.0 |
| 203 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 336 | 1 325–1 347 | 2 781 | Proprietary |
| 204 | Qwen-Plus | Alibaba Qwen | 1 326 | 1 307–1 345 | 732 | Proprietary |
| 205 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 319 | 1 309–1 329 | 3 338 | Proprietary |
| 206 | Step-1o Turbo | StepFun | 1 318 | 1 294–1 342 | 559 | Proprietary |
| 207 | GPT-5 Nano в режиме high | OpenAI | 1 317 | 1 291–1 344 | 492 | Proprietary |
| 208 | gpt-oss-20b | OpenAI | 1 317 | 1 295–1 339 | 672 | Apache 2.0 |
| 209 | OLMo 3 32B Think | Ai2 | 1 315 | 1 282–1 347 | 306 | Apache 2.0 |
| 210 | Gemini 1.5 Pro · 1.5-pro-002 | 1 315 | 1 307–1 322 | 7 610 | Proprietary | |
| 211 | Granite 4.1 8B | IBM Granite | 1 315 | 1 276–1 354 | 236 | Apache 2.0 |
| 212 | Gemma 3 27B | 1 311 | 1 302–1 321 | 3 552 | Gemma | |
| 213 | DeepSeek V3 | DeepSeek | 1 311 | 1 300–1 321 | 2 721 | DeepSeek |
| 214 | Gemini 2.0 Flash-Lite | 1 309 | 1 299–1 319 | 2 814 | Proprietary | |
| 215 | OLMo 3.1 32B Instruct | Ai2 | 1 308 | 1 285–1 331 | 653 | Apache 2.0 |
| 216 | Gemma 3 12B | 1 307 | 1 280–1 334 | 389 | Gemma | |
| 217 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 307 | 1 300–1 313 | 9 999 | Proprietary |
| 218 | Step-2 16k | StepFun | 1 304 | 1 285–1 324 | 642 | Proprietary |
| 219 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 303 | 1 296–1 311 | 11 359 | Proprietary |
| 220 | Athene V2 Chat | Nexusflow | 1 300 | 1 291–1 310 | 3 412 | NexusFlow |
| 221 | Yi-Lightning | 01.AI | 1 300 | 1 290–1 309 | 3 921 | Proprietary |
| 222 | Llama 4 Maverick | Meta | 1 299 | 1 288–1 310 | 2 810 | Llama 4 |
| 223 | Command A | Cohere | 1 299 | 1 289–1 308 | 3 963 | CC-BY-NC-4.0 |
| 224 | Qwen2.5 Plus | Alibaba Qwen | 1 298 | 1 284–1 311 | 1 404 | Proprietary |
| 225 | OLMo 3.1 32B Think | Ai2 | 1 297 | 1 271–1 324 | 453 | Apache 2.0 |
| 226 | Hunyuan TurboS · 20250226 | Tencent | 1 294 | 1 263–1 325 | 238 | Proprietary |
| 227 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 288 | 1 272–1 305 | 1 031 | DeepSeek |
| 228 | GLM-4-Plus · plus-0111 | Z.ai | 1 287 | 1 268–1 307 | 721 | Proprietary |
| 229 | Llama 4 Scout | Meta | 1 286 | 1 272–1 299 | 1 928 | Llama |
| 230 | GPT-4o (2024-08-06) | OpenAI | 1 285 | 1 277–1 293 | 6 826 | Proprietary |
| 231 | GPT-4o (2024-05-13) | OpenAI | 1 284 | 1 278–1 291 | 15 103 | Proprietary |
| 232 | Grok 2 | xAI | 1 283 | 1 276–1 290 | 8 950 | Proprietary |
| 233 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 283 | 1 274–1 291 | 5 415 | Qwen |
| 234 | Llama 3.1 405B Instruct · fp8 | Meta | 1 281 | 1 274–1 289 | 8 482 | Llama 3.1 Community |
| 235 | Hunyuan Large | Tencent | 1 281 | 1 257–1 305 | 497 | Proprietary |
| 236 | Llama 3.1 405B Instruct · bf16 | Meta | 1 278 | 1 270–1 286 | 5 215 | Llama 3.1 Community |
| 237 | Qwen Max | Alibaba Qwen | 1 275 | 1 263–1 287 | 2 249 | Qwen |
| 238 | GLM-4-Plus · plus | Z.ai | 1 275 | 1 265–1 285 | 3 599 | Proprietary |
| 239 | Hunyuan Standard · 2025-02-10 | Tencent | 1 274 | 1 250–1 298 | 499 | Proprietary |
| 240 | GPT-4.1 Nano | OpenAI | 1 274 | 1 251–1 297 | 582 | Proprietary |
| 241 | Hunyuan Turbo | Tencent | 1 273 | 1 242–1 303 | 243 | Proprietary |
| 242 | Claude 3 Opus | Anthropic | 1 273 | 1 266–1 279 | 25 769 | Proprietary |
| 243 | Gemini 1.5 Pro · advanced-0514 | 1 272 | 1 262–1 281 | 6 395 | Proprietary | |
| 244 | GPT-4 Turbo | OpenAI | 1 272 | 1 264–1 279 | 13 217 | Proprietary |
| 245 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 271 | 1 254–1 288 | 1 041 | Llama 3.1 |
| 246 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 271 | 1 261–1 280 | 3 649 | DeepSeek |
| 247 | Gemini 1.5 Pro · 1.5-pro-001 | 1 269 | 1 262–1 277 | 10 492 | Proprietary | |
| 248 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 269 | 1 261–1 277 | 13 306 | Proprietary |
| 249 | Gemini 1.5 Flash · 002 | 1 269 | 1 261–1 278 | 4 789 | Proprietary | |
| 250 | Hunyuan Large Vision | Tencent | 1 268 | 1 238–1 299 | 344 | Proprietary |
| 251 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 268 | 1 260–1 276 | 12 374 | Proprietary |
| 252 | Llama 3.3 70B Instruct | Meta | 1 268 | 1 260–1 275 | 5 767 | Llama-3.3 |
| 253 | GPT-4o-mini (2024-07-18) | OpenAI | 1 267 | 1 260–1 274 | 9 319 | Proprietary |
| 254 | Grok 2 Mini | xAI | 1 265 | 1 257–1 273 | 7 261 | Proprietary |
| 255 | Mistral Small 3.1 24B | Mistral AI | 1 262 | 1 249–1 275 | 2 093 | Apache 2.0 |
| 256 | Mistral Large 2407 | Mistral AI | 1 261 | 1 253–1 270 | 6 664 | Mistral Research |
| 257 | Mistral Large · 2411 | Mistral AI | 1 261 | 1 252–1 270 | 3 574 | MRL |
| 258 | Granite 4.0 H Small | IBM Granite | 1 252 | 1 220–1 285 | 335 | Apache 2.0 |
| 259 | Llama 3.1 70B Instruct | Meta | 1 252 | 1 244–1 259 | 7 677 | Llama 3.1 Community |
| 260 | Nova Pro 1.0 | Amazon | 1 252 | 1 242–1 261 | 2 978 | Proprietary |
| 261 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 251 | 1 231–1 270 | 725 | Apache 2.0 |
| 262 | Gemma 3n E4B | 1 250 | 1 235–1 265 | 1 554 | Gemma | |
| 263 | Magistral Medium | Mistral AI | 1 248 | 1 222–1 274 | 550 | Proprietary |
| 264 | Phi 4 | Microsoft | 1 246 | 1 235–1 256 | 2 764 | MIT |
| 265 | Claude 3.5 Haiku | Anthropic | 1 244 | 1 237–1 252 | 6 319 | Proprietary |
| 266 | Llama 3.1 Tulu 3 70B | Ai2 | 1 242 | 1 217–1 266 | 397 | Llama 3.1 |
| 267 | DeepSeek Coder V2 | DeepSeek | 1 241 | 1 228–1 255 | 1 858 | DeepSeek License |
| 268 | Mistral Small 3 | Mistral AI | 1 240 | 1 227–1 253 | 1 683 | Apache 2.0 |
| 269 | Gemma 3 4B | 1 239 | 1 211–1 267 | 423 | Gemma | |
| 270 | Qwen2 72B Instruct | Alibaba Qwen | 1 235 | 1 225–1 244 | 4 835 | Qianwen LICENSE |
| 271 | Hunyuan Standard · 256k | Tencent | 1 235 | 1 206–1 263 | 361 | Proprietary |
| 272 | Athene 70B | Nexusflow | 1 231 | 1 221–1 242 | 2 921 | CC-BY-NC-4.0 |
| 273 | GPT-4 · 0314 | OpenAI | 1 230 | 1 220–1 240 | 7 052 | Proprietary |
| 274 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 230 | 1 207–1 252 | 507 | Llama 3.1 |
| 275 | Gemini 1.5 Flash · 001 | 1 229 | 1 221–1 237 | 8 392 | Proprietary | |
| 276 | Nova Lite 1.0 | Amazon | 1 227 | 1 216–1 237 | 2 511 | Proprietary |
| 277 | Reka Core | Reka AI | 1 222 | 1 207–1 236 | 1 207 | Proprietary |
| 278 | Jamba 1.5 Large | AI21 Labs | 1 221 | 1 206–1 236 | 1 147 | Jamba Open |
| 279 | GLM-4 | Z.ai | 1 218 | 1 202–1 233 | 1 191 | Proprietary |
| 280 | Llama 3 70B Instruct | Meta | 1 218 | 1 210–1 225 | 20 941 | Llama 3 Community |
| 281 | GPT-4 · 0613 | OpenAI | 1 217 | 1 209–1 225 | 11 181 | Proprietary |
| 282 | Nemotron-4 340B Instruct | NVIDIA | 1 216 | 1 204–1 228 | 2 352 | NVIDIA Open Model |
| 283 | QwQ 32B · 32b-preview | Alibaba Qwen | 1 213 | 1 189–1 238 | 480 | Apache 2.0 |
| 284 | Claude 3 Sonnet | Anthropic | 1 213 | 1 205–1 221 | 13 766 | Proprietary |
| 285 | Gemma 2 27B | 1 212 | 1 205–1 219 | 10 170 | Gemma license | |
| 286 | OLMo 2 32B Instruct | Ai2 | 1 207 | 1 179–1 236 | 375 | Apache-2.0 |
| 287 | Gemini 1.5 Flash-8B | 1 207 | 1 198–1 215 | 5 036 | Proprietary | |
| 288 | Nova Micro 1.0 | Amazon | 1 206 | 1 195–1 217 | 2 455 | Proprietary |
| 289 | Mistral Large · 2402 | Mistral AI | 1 200 | 1 191–1 209 | 7 987 | Proprietary |
| 290 | Aya Expanse 32B | Cohere | 1 200 | 1 190–1 209 | 3 854 | CC-BY-NC-4.0 |
| 291 | Reka Flash (2024-09) | Reka AI | 1 195 | 1 181–1 209 | 1 284 | Proprietary |
| 292 | Llama 3.1 Tulu 3 8B | Ai2 | 1 195 | 1 169–1 220 | 363 | Llama 3.1 |
| 293 | Ministral 8B (2410) | Mistral AI | 1 188 | 1 169–1 208 | 683 | MRL |
| 294 | Claude 3 Haiku | Anthropic | 1 188 | 1 181–1 195 | 14 983 | Proprietary |
| 295 | Command R+ (08-2024) | Cohere | 1 188 | 1 174–1 202 | 1 467 | CC-BY-NC-4.0 |
| 296 | Qwen1.5 110B Chat | Alibaba Qwen | 1 185 | 1 174–1 197 | 3 188 | Qianwen LICENSE |
| 297 | Mixtral 8x22B Instruct | Mistral AI | 1 184 | 1 175–1 193 | 6 778 | Apache 2.0 |
| 298 | Gemma 2 9B | 1 183 | 1 176–1 191 | 7 110 | Gemma license | |
| 299 | Yi-1.5 34B Chat | 01.AI | 1 182 | 1 171–1 193 | 2 985 | Apache-2.0 |
| 300 | Mistral Medium (2023) | Mistral AI | 1 180 | 1 169–1 191 | 4 406 | Proprietary |
Данные на 13 сентября 2026. Как мы считаем: методика
Свежесть данных
Снимок замеров опубликован 13 сентября 2026; всего снимков борда в архиве: 200.
- замеры оценок людей: загружено 18 сентября 2026, 08:25 UTC
- каталог моделей и цены: загружено 18 сентября 2026, 08:51 UTC
- доли использования: загружено 18 сентября 2026, 08:51 UTC
Другие рейтинги
- Рейтинг агентов
- Рейтинг работы с документами
- Рейтинг редактирования изображений
- Рейтинг видео из изображения
- Популярность моделей по использованию через API-агрегаторы
- Рейтинг поисковых моделей
- Рейтинг текстовых моделей
- Рейтинг моделей для кода
- Рейтинг моделей для творческих текстов
- Рейтинг моделей для русского языка
- Рейтинг генерации изображений
- Рейтинг генерации видео
- Рейтинг редактирования видео
- Рейтинг моделей с изображениями на входе
- Рейтинг моделей для веб-разработки
Данные и источники
Обновлено 18 сентября 2026. Каталог, цены и режимы проверяются ежедневно по нескольким API-агрегаторам и провайдерам. Места — по открытым замерам оценок людей, интервалы приведены как в замере.