Рейтинг текстовых моделей
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 «китайский», 300 строк. Оценка и интервал — из источника; интервал не равен диапазону мест.
| Место | Модель | Разработчик | Оценка | Интервал | Голоса | Лицензия |
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5.1 в режиме max | Anthropic | 1 605 | 1 573–1 637 | 369 | Proprietary |
| 2 | Claude Opus 5 в режиме max | Anthropic | 1 585 | 1 569–1 602 | 1 472 | Proprietary |
| 3 | Claude Opus 5 в режиме high | Anthropic | 1 577 | 1 565–1 589 | 3 020 | Proprietary |
| 4 | Gemini 3.7 Flash в режиме high | 1 563 | 1 532–1 595 | 386 | Proprietary | |
| 5 | Gemini 3.8 Flash в режиме high | 1 556 | 1 524–1 588 | 358 | Proprietary | |
| 6 | Claude Opus 4.6 в режиме high | Anthropic | 1 551 | 1 541–1 561 | 4 483 | Proprietary |
| 7 | Claude Opus 4.6 | Anthropic | 1 547 | 1 538–1 557 | 4 515 | Proprietary |
| 8 | Claude Fable 5 | Anthropic | 1 542 | 1 528–1 557 | 1 873 | Proprietary |
| 9 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 539 | 1 520–1 557 | 1 097 | Proprietary |
| 10 | Muse Spark 1.3 в режиме max | Meta | 1 538 | 1 505–1 572 | 299 | Proprietary |
| 11 | GPT-5.5 · 5.5 | OpenAI | 1 537 | 1 526–1 548 | 3 656 | Proprietary |
| 12 | Gemini 3.6 Flash в режиме high | 1 535 | 1 520–1 549 | 1 717 | Proprietary | |
| 13 | Claude Opus 4.7 в режиме high | Anthropic | 1 534 | 1 523–1 545 | 3 369 | Proprietary |
| 14 | Qwen3.5 Max | Alibaba Qwen | 1 532 | 1 516–1 548 | 1 391 | Proprietary |
| 15 | Gemini 3.1 Pro Preview | 1 530 | 1 522–1 538 | 6 460 | Proprietary | |
| 16 | Gemini 3.5 Flash в режиме high | 1 529 | 1 517–1 542 | 2 514 | Proprietary | |
| 17 | Kimi K2.6 | Moonshot AI | 1 529 | 1 515–1 543 | 1 886 | Modified MIT |
| 18 | GLM 5.3 Flash | Z.ai | 1 527 | 1 502–1 552 | 602 | MIT |
| 19 | Claude Opus 4.7 | Anthropic | 1 525 | 1 514–1 536 | 3 433 | Proprietary |
| 20 | Gemini 3 Pro | 1 525 | 1 514–1 536 | 3 172 | Proprietary | |
| 21 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 525 | 1 510–1 539 | 1 829 | Proprietary |
| 22 | Qwen3.7 Max | Alibaba Qwen | 1 523 | 1 486–1 560 | 265 | Proprietary |
| 23 | GPT-5.5 в режиме high | OpenAI | 1 522 | 1 511–1 532 | 3 770 | Proprietary |
| 24 | Gemini 3.5 Flash в режиме medium | 1 521 | 1 508–1 534 | 2 273 | Proprietary | |
| 25 | Muse Spark 1.2 в режиме xhigh | Meta | 1 521 | 1 483–1 559 | 256 | Proprietary |
| 26 | Kimi K3 в режиме max | Moonshot AI | 1 521 | 1 504–1 537 | 1 386 | Kimi K3 license |
| 27 | Muse Spark 1.1 | Meta | 1 519 | 1 505–1 534 | 1 911 | Proprietary |
| 28 | GPT-5.4 в режиме high | OpenAI | 1 517 | 1 506–1 528 | 3 462 | Proprietary |
| 29 | GLM 5.1 | Z.ai | 1 517 | 1 505–1 528 | 2 984 | MIT |
| 30 | GLM 5.2 в режиме max | Z.ai | 1 516 | 1 503–1 529 | 2 282 | MIT |
| 31 | GLM 5.3 в режиме max | Z.ai | 1 516 | 1 494–1 539 | 750 | MIT |
| 32 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 514 | 1 500–1 528 | 1 918 | Proprietary |
| 33 | GLM 5 | Z.ai | 1 511 | 1 496–1 525 | 1 733 | MIT |
| 34 | Qwen3.7 Plus | Alibaba Qwen | 1 510 | 1 497–1 523 | 2 216 | Proprietary |
| 35 | ERNIE 5.0 · preview-1022 | Baidu | 1 509 | 1 479–1 540 | 381 | Proprietary |
| 36 | Gemini 2.5 Pro | 1 509 | 1 502–1 516 | 8 792 | Proprietary | |
| 37 | Gemini 3 Flash Preview | 1 509 | 1 496–1 521 | 2 256 | Proprietary | |
| 38 | GPT-5.4 | OpenAI | 1 508 | 1 498–1 519 | 3 675 | Proprietary |
| 39 | MiMo-V2.5-Pro | Xiaomi | 1 506 | 1 495–1 517 | 3 376 | MIT |
| 40 | Claude Opus 4.8 | Anthropic | 1 506 | 1 494–1 517 | 3 128 | Proprietary |
| 41 | Muse Spark | Meta | 1 504 | 1 483–1 525 | 849 | Proprietary |
| 42 | Grok 4.5 | xAI | 1 502 | 1 488–1 516 | 2 028 | Proprietary |
| 43 | Hy3 | Tencent | 1 501 | 1 476–1 527 | 546 | Apache 2.0 |
| 44 | ERNIE 5.1 | Baidu | 1 501 | 1 487–1 516 | 1 736 | Proprietary |
| 45 | Qwen3.8 27B | Alibaba Qwen | 1 500 | 1 477–1 522 | 686 | Apache 2.0 |
| 46 | GLM 4.6 | Z.ai | 1 500 | 1 488–1 511 | 2 590 | MIT |
| 47 | Claude Opus 4.8 в режиме high | Anthropic | 1 499 | 1 488–1 511 | 3 114 | Proprietary |
| 48 | ERNIE 5.0 · 0110 | Baidu | 1 499 | 1 487–1 512 | 2 237 | Proprietary |
| 49 | Kimi K2.5 · thinking | Moonshot AI | 1 498 | 1 488–1 508 | 3 933 | Modified MIT |
| 50 | ERNIE 5.0 · preview-1203 | Baidu | 1 497 | 1 476–1 518 | 792 | Proprietary |
| 51 | Qwen3.5 397B A17B | Alibaba Qwen | 1 495 | 1 486–1 505 | 4 604 | Apache 2.0 |
| 52 | GLM 4.7 | Z.ai | 1 495 | 1 477–1 514 | 965 | MIT |
| 53 | Grok 4.6 в режиме high | xAI | 1 494 | 1 476–1 513 | 1 110 | Proprietary |
| 54 | GPT-5.1 в режиме high | OpenAI | 1 494 | 1 483–1 504 | 3 123 | Proprietary |
| 55 | GLM 5V Turbo | Z.ai | 1 492 | 1 468–1 516 | 602 | Proprietary |
| 56 | Nemotron 3 Ultra | NVIDIA | 1 491 | 1 467–1 514 | 627 | OpenMDW-1.1 |
| 57 | Inkling | Thinking Machines Lab | 1 490 | 1 476–1 504 | 1 820 | Apache 2.0 |
| 58 | Seed 2.0 Pro | ByteDance Seed | 1 489 | 1 480–1 499 | 4 372 | Proprietary |
| 59 | Claude Sonnet 4.6 | Anthropic | 1 489 | 1 479–1 499 | 3 742 | Proprietary |
| 60 | Qwen3.6 Max Preview | Alibaba Qwen | 1 489 | 1 456–1 522 | 319 | Proprietary |
| 61 | DeepSeek V4 Pro 0423 | DeepSeek | 1 485 | 1 474–1 497 | 2 859 | MIT |
| 62 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 483 | 1 457–1 509 | 539 | Apache 2.0 |
| 63 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 483 | 1 471–1 495 | 2 747 | MIT |
| 64 | Gemma 4 26B A4B | 1 483 | 1 457–1 509 | 469 | Apache 2.0 | |
| 65 | Grok 4.20 · beta-0309-reasoning | xAI | 1 481 | 1 470–1 491 | 3 572 | Proprietary |
| 66 | Qwen3 Max · preview | Alibaba Qwen | 1 479 | 1 466–1 492 | 2 092 | Proprietary |
| 67 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 479 | 1 454–1 504 | 585 | MIT |
| 68 | Claude Sonnet 5 в режиме high | Anthropic | 1 478 | 1 466–1 491 | 2 361 | Proprietary |
| 69 | Grok 4.20 Multi-Agent | xAI | 1 478 | 1 467–1 489 | 3 391 | Proprietary |
| 70 | Gemini 3 Flash Preview в режиме minimal | 1 477 | 1 468–1 486 | 5 067 | Proprietary | |
| 71 | Grok 4.20 · beta1 | xAI | 1 477 | 1 461–1 492 | 1 540 | Proprietary |
| 72 | Qwen3.6 Plus | Alibaba Qwen | 1 472 | 1 459–1 485 | 2 219 | Proprietary |
| 73 | Qwen3.5-27B | Alibaba Qwen | 1 472 | 1 458–1 486 | 1 750 | Apache 2.0 |
| 74 | Gemma 4 31B | 1 472 | 1 445–1 498 | 465 | Apache 2.0 | |
| 75 | MiMo-V2.5 | Xiaomi | 1 471 | 1 457–1 484 | 2 136 | MIT |
| 76 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 470 | 1 452–1 488 | 1 150 | MIT |
| 77 | GPT-5.1 | OpenAI | 1 470 | 1 459–1 480 | 3 288 | Proprietary |
| 78 | Claude Opus 4.5 | Anthropic | 1 469 | 1 460–1 478 | 4 791 | Proprietary |
| 79 | MiMo-V2 Omni | Xiaomi | 1 468 | 1 448–1 488 | 856 | Proprietary |
| 80 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 468 | 1 455–1 480 | 2 501 | MIT |
| 81 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 467 | 1 453–1 481 | 1 958 | Proprietary |
| 82 | Gemini 3.5 Flash Lite | 1 466 | 1 452–1 481 | 1 812 | Proprietary | |
| 83 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 466 | 1 451–1 481 | 1 722 | Proprietary |
| 84 | GLM 4.5 | Z.ai | 1 466 | 1 452–1 480 | 1 786 | MIT |
| 85 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 466 | 1 452–1 480 | 1 863 | Apache 2.0 |
| 86 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 466 | 1 452–1 480 | 1 797 | Proprietary |
| 87 | Grok 4.1 · 4.1 | xAI | 1 465 | 1 456–1 474 | 4 842 | Proprietary |
| 88 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 465 | 1 458–1 473 | 6 826 | Apache 2.0 |
| 89 | DeepSeek V4 Flash 0423 | DeepSeek | 1 465 | 1 453–1 477 | 2 526 | MIT |
| 90 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 465 | 1 454–1 476 | 3 014 | MIT |
| 91 | Claude Opus 4.5 в режиме high | Anthropic | 1 465 | 1 453–1 476 | 2 755 | Proprietary |
| 92 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 463 | 1 444–1 481 | 998 | MIT |
| 93 | Gemini 3.1 Flash Lite Preview | 1 462 | 1 452–1 473 | 3 578 | Proprietary | |
| 94 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 462 | 1 438–1 485 | 698 | Apache 2.0 |
| 95 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 462 | 1 447–1 476 | 1 711 | Apache 2.0 |
| 96 | GPT-5.2 Chat | OpenAI | 1 461 | 1 448–1 474 | 2 100 | Proprietary |
| 97 | MiniMax M3 | MiniMax | 1 461 | 1 448–1 473 | 2 640 | MiniMax Community License |
| 98 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 460 | 1 446–1 474 | 1 825 | Apache 2.0 |
| 99 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 460 | 1 434–1 486 | 498 | MIT |
| 100 | Claude Sonnet 4.5 | Anthropic | 1 459 | 1 450–1 468 | 5 298 | Proprietary |
| 101 | Kimi K2.5 · instant | Moonshot AI | 1 459 | 1 432–1 486 | 448 | Modified MIT |
| 102 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 458 | 1 449–1 466 | 5 298 | Proprietary |
| 103 | Grok 4.1 · 4.1-thinking | xAI | 1 458 | 1 448–1 467 | 4 572 | Proprietary |
| 104 | GPT-5.5 · 5.5-instant | OpenAI | 1 457 | 1 441–1 474 | 1 487 | Proprietary |
| 105 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 457 | 1 424–1 490 | 291 | Proprietary |
| 106 | Grok 4 Fast · chat | xAI | 1 457 | 1 428–1 485 | 461 | Proprietary |
| 107 | ChatGPT-4o (latest) | OpenAI | 1 456 | 1 448–1 464 | 5 633 | Proprietary |
| 108 | Kimi K2 Thinking | Moonshot AI | 1 456 | 1 446–1 465 | 4 264 | Modified MIT |
| 109 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 455 | 1 445–1 466 | 3 413 | MIT |
| 110 | MiMo-V2 Pro | Xiaomi | 1 455 | 1 439–1 472 | 1 408 | Proprietary |
| 111 | Inkling Small | Thinking Machines Lab | 1 455 | 1 438–1 471 | 1 351 | Apache 2.0 |
| 112 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 453 | 1 437–1 468 | 1 407 | MIT |
| 113 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 452 | 1 441–1 463 | 2 809 | MIT |
| 114 | Qwen3.5-Flash | Alibaba Qwen | 1 450 | 1 439–1 461 | 3 146 | Proprietary |
| 115 | Gemini 2.5 Flash · flash | 1 450 | 1 443–1 457 | 8 638 | Proprietary | |
| 116 | GPT-5.2 в режиме high | OpenAI | 1 450 | 1 439–1 460 | 3 136 | Proprietary |
| 117 | Mistral Medium 3.1 | Mistral AI | 1 450 | 1 442–1 457 | 6 637 | Proprietary |
| 118 | Step 3.5 Flash | StepFun | 1 449 | 1 438–1 460 | 3 117 | Apache 2.0 |
| 119 | LongCat-Flash-Chat · chat | Meituan | 1 449 | 1 429–1 469 | 869 | MIT |
| 120 | GPT-5.4 Mini в режиме high | OpenAI | 1 449 | 1 438–1 460 | 3 298 | Proprietary |
| 121 | Grok 3 | xAI | 1 449 | 1 435–1 462 | 1 999 | Proprietary |
| 122 | GPT-5.2 | OpenAI | 1 448 | 1 439–1 458 | 4 528 | Proprietary |
| 123 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 448 | 1 436–1 460 | 2 269 | Proprietary | |
| 124 | Mistral Large 3 2512 | Mistral AI | 1 446 | 1 437–1 455 | 4 738 | Apache 2.0 |
| 125 | Mistral Medium 3.5 | Mistral AI | 1 445 | 1 419–1 471 | 502 | Modified MIT |
| 126 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 444 | 1 415–1 472 | 433 | Apache 2.0 |
| 127 | R1 0528 | DeepSeek | 1 444 | 1 426–1 461 | 1 188 | MIT |
| 128 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 443 | 1 412–1 473 | 314 | Proprietary |
| 129 | Grok 4.1 Fast | xAI | 1 443 | 1 433–1 452 | 3 938 | Proprietary |
| 130 | Hy3 preview | Tencent | 1 441 | 1 411–1 472 | 401 | tencent-hunyuan-community |
| 131 | o3 | OpenAI | 1 438 | 1 428–1 448 | 3 866 | Proprietary |
| 132 | Grok 4 Fast · reasoning | xAI | 1 438 | 1 420–1 455 | 1 093 | Proprietary |
| 133 | MiniMax M2.7 | MiniMax | 1 435 | 1 425–1 445 | 3 914 | Modified MIT |
| 134 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 433 | 1 419–1 448 | 1 746 | Apache 2.0 |
| 135 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 433 | 1 408–1 458 | 495 | Proprietary |
| 136 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 433 | 1 411–1 454 | 698 | MIT |
| 137 | GPT-5.3 Chat | OpenAI | 1 430 | 1 417–1 444 | 1 978 | Proprietary |
| 138 | Muse Glimmer 30B | Meta | 1 430 | 1 396–1 464 | 297 | Apache-2.0 |
| 139 | MiniMax M2.1 | MiniMax | 1 429 | 1 412–1 445 | 1 255 | MIT |
| 140 | Grok 4 | xAI | 1 428 | 1 417–1 440 | 2 781 | Proprietary |
| 141 | Nemotron 3 Super | NVIDIA | 1 428 | 1 402–1 453 | 500 | NVIDIA Open Model |
| 142 | Claude Opus 4.1 · 20250805 | Anthropic | 1 427 | 1 419–1 436 | 5 665 | Proprietary |
| 143 | GLM 4.6V | Z.ai | 1 427 | 1 392–1 463 | 260 | MIT |
| 144 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 426 | 1 416–1 436 | 3 738 | Proprietary |
| 145 | Hunyuan T1 | Tencent | 1 426 | 1 395–1 458 | 353 | Proprietary |
| 146 | GLM 4.5 Air | Z.ai | 1 426 | 1 413–1 439 | 2 218 | MIT |
| 147 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 426 | 1 414–1 438 | 2 583 | Apache 2.0 |
| 148 | GPT-5 | OpenAI | 1 424 | 1 412–1 436 | 2 324 | Proprietary |
| 149 | Grok 4.3 | xAI | 1 423 | 1 413–1 434 | 3 775 | Proprietary |
| 150 | GPT-5 в режиме high | OpenAI | 1 423 | 1 410–1 436 | 2 223 | Proprietary |
| 151 | GPT-4.5 Preview | OpenAI | 1 421 | 1 402–1 441 | 907 | Proprietary |
| 152 | Ling-flash-2.0 | inclusionAI | 1 421 | 1 393–1 449 | 433 | MIT |
| 153 | Claude Haiku 4.5 | Anthropic | 1 416 | 1 409–1 423 | 8 321 | Proprietary |
| 154 | Hunyuan TurboS · 20250416 | Tencent | 1 416 | 1 391–1 441 | 539 | Proprietary |
| 155 | Kimi K2 0905 | Moonshot AI | 1 416 | 1 397–1 435 | 948 | Modified MIT |
| 156 | GLM 4.7 Flash | Z.ai | 1 415 | 1 393–1 437 | 700 | MIT |
| 157 | Granite 4.2 30B | IBM Granite | 1 414 | 1 373–1 455 | 201 | Apache 2.0 |
| 158 | Mercury 2 | Inception Labs | 1 412 | 1 372–1 452 | 200 | Proprietary |
| 159 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 406 | 1 388–1 424 | 1 042 | Apache 2.0 |
| 160 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 403 | 1 393–1 413 | 3 415 | Proprietary | |
| 161 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 403 | 1 383–1 422 | 940 | Proprietary |
| 162 | Kimi K2 0711 | Moonshot AI | 1 402 | 1 389–1 415 | 2 084 | Modified MIT |
| 163 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 401 | 1 389–1 414 | 2 252 | Proprietary | |
| 164 | R1 | DeepSeek | 1 400 | 1 382–1 418 | 1 151 | MIT |
| 165 | Step 3 | StepFun | 1 398 | 1 366–1 431 | 308 | Apache 2.0 |
| 166 | Ring-flash-2.0 | inclusionAI | 1 396 | 1 369–1 424 | 422 | MIT |
| 167 | MiniMax M2.5 | MiniMax | 1 394 | 1 382–1 406 | 2 482 | Modified MIT |
| 168 | o1 · 2024-12-17 | OpenAI | 1 394 | 1 380–1 408 | 1 780 | Proprietary |
| 169 | DeepSeek V3 0324 | DeepSeek | 1 390 | 1 380–1 401 | 3 031 | MIT |
| 170 | GPT-5.4 Nano в режиме high | OpenAI | 1 389 | 1 378–1 401 | 3 194 | Proprietary |
| 171 | GLM-4-Plus · plus-0111 | Z.ai | 1 389 | 1 359–1 419 | 387 | Proprietary |
| 172 | GPT-5 Mini в режиме high | OpenAI | 1 388 | 1 374–1 401 | 1 922 | Proprietary |
| 173 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 386 | 1 370–1 403 | 1 251 | NVIDIA Open Model |
| 174 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 386 | 1 373–1 398 | 2 430 | Proprietary |
| 175 | Grok 3 Mini в режиме high | xAI | 1 385 | 1 365–1 404 | 871 | Proprietary |
| 176 | gpt-oss-120b | OpenAI | 1 384 | 1 372–1 397 | 2 245 | Apache 2.0 |
| 177 | Claude Opus 4 · 20250514 | Anthropic | 1 383 | 1 372–1 394 | 3 106 | Proprietary |
| 178 | Qwen2.5 Max | Alibaba Qwen | 1 381 | 1 368–1 394 | 2 074 | Proprietary |
| 179 | GPT-4.1 | OpenAI | 1 381 | 1 370–1 392 | 3 277 | Proprietary |
| 180 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 379 | 1 365–1 394 | 1 533 | Apache 2.0 |
| 181 | QwQ 32B · 32b | Alibaba Qwen | 1 379 | 1 364–1 394 | 1 530 | Apache 2.0 |
| 182 | Step-1o Turbo | StepFun | 1 379 | 1 354–1 405 | 538 | Proprietary |
| 183 | o3 Mini High | OpenAI | 1 379 | 1 361–1 396 | 1 181 | Proprietary |
| 184 | Grok 3 Mini | xAI | 1 377 | 1 361–1 394 | 1 304 | Proprietary |
| 185 | Mistral Medium 3 | Mistral AI | 1 373 | 1 360–1 386 | 2 043 | Proprietary |
| 186 | Gemini 2.0 Flash | 1 373 | 1 361–1 384 | 2 692 | Proprietary | |
| 187 | MiniMax M2 | MiniMax | 1 371 | 1 343–1 398 | 447 | Apache 2.0 |
| 188 | Trinity Large Thinking | Arcee AI | 1 370 | 1 355–1 386 | 1 817 | Apache 2.0 |
| 189 | Nova 2 Lite | Amazon | 1 365 | 1 347–1 383 | 1 031 | Proprietary |
| 190 | INTELLECT-3 | Prime Intellect | 1 362 | 1 335–1 389 | 482 | MIT |
| 191 | MiniMax M1 | MiniMax | 1 360 | 1 348–1 372 | 2 340 | Apache 2.0 |
| 192 | Hunyuan Turbo | Tencent | 1 358 | 1 318–1 397 | 208 | Proprietary |
| 193 | GPT-5 Nano в режиме high | OpenAI | 1 357 | 1 331–1 383 | 512 | Proprietary |
| 194 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 356 | 1 341–1 370 | 1 789 | Apache 2.0 |
| 195 | Qwen3 32B | Alibaba Qwen | 1 355 | 1 318–1 392 | 189 | Apache 2.0 |
| 196 | o4 Mini | OpenAI | 1 354 | 1 343–1 365 | 3 017 | Proprietary |
| 197 | Trinity Large | Arcee AI | 1 351 | 1 336–1 366 | 1 834 | Apache 2.0 |
| 198 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 350 | 1 338–1 363 | 2 358 | Proprietary |
| 199 | Qwen-Plus | Alibaba Qwen | 1 347 | 1 319–1 374 | 459 | Proprietary |
| 200 | Qwen3 30B A3B | Alibaba Qwen | 1 346 | 1 331–1 361 | 1 577 | Apache 2.0 |
| 201 | Gemma 3 27B | 1 346 | 1 335–1 357 | 2 830 | Gemma | |
| 202 | Hunyuan Large | Tencent | 1 346 | 1 313–1 378 | 295 | Proprietary |
| 203 | Amazon Nova Experimental Chat · 10-09 | Amazon | 1 343 | 1 305–1 381 | 217 | Proprietary |
| 204 | Mistral Small 3.2 24B | Mistral AI | 1 342 | 1 325–1 359 | 1 089 | Apache 2.0 |
| 205 | Claude Sonnet 4 · 20250514 | Anthropic | 1 341 | 1 329–1 353 | 2 841 | Proprietary |
| 206 | Gemini 2.0 Flash-Lite | 1 338 | 1 324–1 353 | 1 509 | Proprietary | |
| 207 | GLM 4.5V | Z.ai | 1 338 | 1 302–1 375 | 234 | MIT |
| 208 | DeepSeek V3 | DeepSeek | 1 338 | 1 323–1 354 | 1 450 | DeepSeek |
| 209 | Nemotron 3.5 Lightning | NVIDIA | 1 335 | 1 306–1 363 | 498 | OpenMDW-1.1 |
| 210 | Gemini 1.5 Pro · 1.5-pro-002 | 1 331 | 1 321–1 341 | 4 444 | Proprietary | |
| 211 | Step-2 16k | StepFun | 1 331 | 1 299–1 363 | 328 | Proprietary |
| 212 | GPT-4.1 Mini | OpenAI | 1 329 | 1 317–1 341 | 2 553 | Proprietary |
| 213 | Command A | Cohere | 1 326 | 1 316–1 336 | 3 815 | CC-BY-NC-4.0 |
| 214 | o1 · preview | OpenAI | 1 325 | 1 313–1 338 | 2 763 | Proprietary |
| 215 | Granite 4.1 8B | IBM Granite | 1 324 | 1 277–1 370 | 209 | Apache 2.0 |
| 216 | Hunyuan TurboS · 20250226 | Tencent | 1 323 | 1 284–1 363 | 203 | Proprietary |
| 217 | Yi-Lightning | 01.AI | 1 320 | 1 307–1 332 | 2 736 | Proprietary |
| 218 | o3 Mini | OpenAI | 1 320 | 1 310–1 330 | 3 884 | Proprietary |
| 219 | Hunyuan Standard · 2025-02-10 | Tencent | 1 319 | 1 288–1 350 | 340 | Proprietary |
| 220 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 318 | 1 291–1 344 | 446 | DeepSeek |
| 221 | o1-mini | OpenAI | 1 314 | 1 304–1 324 | 4 309 | Proprietary |
| 222 | gpt-oss-20b | OpenAI | 1 314 | 1 291–1 337 | 662 | Apache 2.0 |
| 223 | Qwen2.5 Plus | Alibaba Qwen | 1 314 | 1 291–1 336 | 621 | Proprietary |
| 224 | Gemma 3n E4B | 1 309 | 1 293–1 324 | 1 491 | Gemma | |
| 225 | OLMo 3 32B Think | Ai2 | 1 304 | 1 275–1 333 | 452 | Apache 2.0 |
| 226 | Athene V2 Chat | Nexusflow | 1 302 | 1 288–1 316 | 1 781 | NexusFlow |
| 227 | OLMo 3.1 32B Instruct | Ai2 | 1 300 | 1 278–1 322 | 792 | Apache 2.0 |
| 228 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 299 | 1 287–1 312 | 2 322 | Proprietary |
| 229 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 297 | 1 285–1 309 | 2 528 | Proprietary |
| 230 | GLM-4-Plus · plus | Z.ai | 1 296 | 1 283–1 309 | 2 427 | Proprietary |
| 231 | Gemini 1.5 Flash · 002 | 1 295 | 1 283–1 306 | 2 980 | Proprietary | |
| 232 | Grok 2 | xAI | 1 289 | 1 280–1 299 | 5 317 | Proprietary |
| 233 | Hunyuan Large Vision | Tencent | 1 288 | 1 256–1 320 | 326 | Proprietary |
| 234 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 281 | 1 268–1 294 | 2 242 | DeepSeek |
| 235 | GPT-4o (2024-05-13) | OpenAI | 1 277 | 1 269–1 286 | 11 416 | Proprietary |
| 236 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 277 | 1 238–1 316 | 211 | Nvidia |
| 237 | Llama 4 Maverick | Meta | 1 277 | 1 264–1 289 | 2 617 | Llama 4 |
| 238 | Gemini 1.5 Pro · 1.5-pro-001 | 1 275 | 1 265–1 285 | 8 912 | Proprietary | |
| 239 | Gemini 1.5 Pro · advanced-0514 | 1 274 | 1 263–1 285 | 5 932 | Proprietary | |
| 240 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 272 | 1 261–1 283 | 3 329 | Qwen |
| 241 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 271 | 1 263–1 280 | 6 916 | Proprietary |
| 242 | GPT-4.1 Nano | OpenAI | 1 269 | 1 234–1 305 | 247 | Proprietary |
| 243 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 265 | 1 255–1 274 | 8 100 | Proprietary |
| 244 | GPT-4o-mini (2024-07-18) | OpenAI | 1 264 | 1 255–1 274 | 5 777 | Proprietary |
| 245 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 263 | 1 242–1 284 | 723 | Llama 3.1 |
| 246 | Grok 2 Mini | xAI | 1 262 | 1 252–1 272 | 4 510 | Proprietary |
| 247 | Llama 4 Scout | Meta | 1 255 | 1 241–1 269 | 2 086 | Llama |
| 248 | GPT-4o (2024-08-06) | OpenAI | 1 254 | 1 243–1 265 | 3 778 | Proprietary |
| 249 | Qwen Max | Alibaba Qwen | 1 254 | 1 238–1 270 | 1 421 | Qwen |
| 250 | Mistral Small 3.1 24B | Mistral AI | 1 252 | 1 238–1 265 | 2 213 | Apache 2.0 |
| 251 | Llama 3.1 Tulu 3 70B | Ai2 | 1 249 | 1 214–1 285 | 224 | Llama 3.1 |
| 252 | Claude 3 Opus | Anthropic | 1 248 | 1 240–1 256 | 23 348 | Proprietary |
| 253 | Nova Pro 1.0 | Amazon | 1 244 | 1 229–1 258 | 1 627 | Proprietary |
| 254 | OLMo 3.1 32B Think | Ai2 | 1 243 | 1 219–1 268 | 697 | Apache 2.0 |
| 255 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 242 | 1 232–1 252 | 9 150 | Proprietary |
| 256 | GPT-4 Turbo | OpenAI | 1 242 | 1 233–1 252 | 10 827 | Proprietary |
| 257 | Llama 3.1 405B Instruct · bf16 | Meta | 1 242 | 1 231–1 253 | 3 105 | Llama 3.1 Community |
| 258 | Hunyuan Standard · 256k | Tencent | 1 241 | 1 207–1 275 | 305 | Proprietary |
| 259 | Reka Core | Reka AI | 1 241 | 1 218–1 264 | 598 | Proprietary |
| 260 | Mistral Large 2407 | Mistral AI | 1 240 | 1 230–1 251 | 3 933 | Mistral Research |
| 261 | Qwen2 72B Instruct | Alibaba Qwen | 1 240 | 1 228–1 251 | 4 078 | Qianwen LICENSE |
| 262 | Granite 4.0 H Small | IBM Granite | 1 239 | 1 206–1 272 | 333 | Apache 2.0 |
| 263 | Mistral Large | Mistral AI | 1 239 | 1 225–1 252 | 1 914 | MRL |
| 264 | Athene 70B | Nexusflow | 1 237 | 1 223–1 252 | 1 678 | CC-BY-NC-4.0 |
| 265 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 237 | 1 228–1 247 | 11 382 | Proprietary |
| 266 | Llama 3.1 405B Instruct · fp8 | Meta | 1 234 | 1 225–1 244 | 5 226 | Llama 3.1 Community |
| 267 | Gemini 1.5 Flash · 001 | 1 234 | 1 224–1 244 | 7 043 | Proprietary | |
| 268 | Gemini 1.5 Flash-8B | 1 231 | 1 219–1 242 | 2 914 | Proprietary | |
| 269 | GLM-4 | Z.ai | 1 228 | 1 212–1 245 | 1 227 | Proprietary |
| 270 | Claude 3.5 Haiku | Anthropic | 1 228 | 1 219–1 238 | 4 667 | Proprietary |
| 271 | Command R+ (08-2024) | Cohere | 1 226 | 1 207–1 246 | 845 | CC-BY-NC-4.0 |
| 272 | Magistral Medium | Mistral AI | 1 225 | 1 201–1 250 | 718 | Proprietary |
| 273 | Nova Lite 1.0 | Amazon | 1 225 | 1 209–1 241 | 1 284 | Proprietary |
| 274 | Gemma 2 9B IT SimPO | Princeton NLP | 1 225 | 1 204–1 246 | 753 | MIT |
| 275 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 222 | 1 198–1 246 | 528 | Apache 2.0 |
| 276 | Gemma 2 27B | 1 221 | 1 212–1 230 | 6 872 | Gemma license | |
| 277 | Reka Flash (2024-09) | Reka AI | 1 221 | 1 198–1 244 | 607 | Proprietary |
| 278 | QwQ 32B · 32b-preview | Alibaba Qwen | 1 218 | 1 177–1 258 | 206 | Apache 2.0 |
| 279 | Llama 3.3 70B Instruct | Meta | 1 217 | 1 207–1 228 | 3 674 | Llama-3.3 |
| 280 | Llama 3.1 70B Instruct | Meta | 1 215 | 1 205–1 225 | 4 846 | Llama 3.1 Community |
| 281 | Nemotron-4 340B Instruct | NVIDIA | 1 214 | 1 200–1 227 | 2 402 | NVIDIA Open Model |
| 282 | Yi-1.5 34B Chat | 01.AI | 1 213 | 1 200–1 226 | 2 868 | Apache-2.0 |
| 283 | Phi 4 | Microsoft | 1 211 | 1 196–1 227 | 1 503 | MIT |
| 284 | Aya Expanse 32B | Cohere | 1 211 | 1 198–1 223 | 2 375 | CC-BY-NC-4.0 |
| 285 | Nova Micro 1.0 | Amazon | 1 209 | 1 193–1 225 | 1 281 | Proprietary |
| 286 | Qwen1.5 110B Chat | Alibaba Qwen | 1 206 | 1 192–1 219 | 2 906 | Qianwen LICENSE |
| 287 | Mistral Small 3 | Mistral AI | 1 204 | 1 186–1 221 | 1 054 | Apache 2.0 |
| 288 | Jamba 1.5 Large | AI21 Labs | 1 203 | 1 181–1 225 | 675 | Jamba Open |
| 289 | DeepSeek Coder V2 | DeepSeek | 1 201 | 1 186–1 216 | 1 712 | DeepSeek License |
| 290 | InternLM2.5 20B Chat | InternLM | 1 199 | 1 181–1 218 | 959 | Other |
| 291 | Ministral 8B (2410) | Mistral AI | 1 193 | 1 168–1 218 | 478 | MRL |
| 292 | OLMo 2 32B Instruct | Ai2 | 1 191 | 1 155–1 227 | 229 | Apache-2.0 |
| 293 | Command R+ | Cohere | 1 191 | 1 181–1 201 | 10 474 | CC-BY-NC-4.0 |
| 294 | Claude 3 Sonnet | Anthropic | 1 189 | 1 179–1 198 | 15 648 | Proprietary |
| 295 | Qwen1.5 72B Chat | Alibaba Qwen | 1 186 | 1 174–1 198 | 4 167 | Qianwen LICENSE |
| 296 | Gemma 2 9B | 1 185 | 1 175–1 195 | 4 980 | Gemma license | |
| 297 | GPT-4 | OpenAI | 1 184 | 1 172–1 196 | 5 849 | Proprietary |
| 298 | Command R (08-2024) | Cohere | 1 182 | 1 162–1 202 | 834 | CC-BY-NC-4.0 |
| 299 | Aya Expanse 8B | Cohere | 1 181 | 1 158–1 204 | 618 | CC-BY-NC-4.0 |
| 300 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 180 | 1 152–1 207 | 403 | Llama 3.1 |
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 моделей для кода
- Рейтинг моделей для творческих текстов
- Рейтинг моделей для математики
- 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.