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 «отрасль: естественные и социальные науки», 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 | 1 531 | 1 520–1 541 | 3 410 | Proprietary |
| 2 | Claude Fable 5.1 в режиме max | Anthropic | 1 526 | 1 506–1 545 | 924 | Proprietary |
| 3 | Claude Opus 5 в режиме high | Anthropic | 1 522 | 1 514–1 530 | 7 274 | Proprietary |
| 4 | Claude Opus 4.6 в режиме high | Anthropic | 1 519 | 1 513–1 526 | 11 837 | Proprietary |
| 5 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 516 | 1 504–1 528 | 2 551 | Proprietary |
| 6 | Muse Spark 1.2 в режиме xhigh | Meta | 1 514 | 1 489–1 538 | 518 | Proprietary |
| 7 | Claude Opus 4.6 | Anthropic | 1 511 | 1 505–1 517 | 12 591 | Proprietary |
| 8 | Gemini 3.7 Flash в режиме high | 1 509 | 1 490–1 529 | 948 | Proprietary | |
| 9 | Gemini 3.8 Flash в режиме high | 1 508 | 1 488–1 528 | 858 | Proprietary | |
| 10 | Claude Fable 5 | Anthropic | 1 508 | 1 499–1 517 | 4 806 | Proprietary |
| 11 | Claude Opus 4.7 в режиме high | Anthropic | 1 508 | 1 501–1 515 | 10 103 | Proprietary |
| 12 | GLM 5.3 в режиме max | Z.ai | 1 504 | 1 490–1 518 | 1 826 | MIT |
| 13 | Muse Spark 1.3 в режиме max | Meta | 1 503 | 1 482–1 524 | 747 | Proprietary |
| 14 | Kimi K3 в режиме max | Moonshot AI | 1 500 | 1 490–1 511 | 3 255 | Kimi K3 license |
| 15 | Gemini 3.1 Pro Preview | 1 499 | 1 494–1 505 | 17 641 | Proprietary | |
| 16 | Qwen3.5 Max | Alibaba Qwen | 1 497 | 1 486–1 507 | 3 486 | Proprietary |
| 17 | Claude Opus 4.7 | Anthropic | 1 495 | 1 488–1 502 | 10 294 | Proprietary |
| 18 | Gemini 3.5 Flash в режиме medium | 1 493 | 1 485–1 502 | 5 904 | Proprietary | |
| 19 | Gemini 3.5 Flash в режиме high | 1 493 | 1 485–1 500 | 6 254 | Proprietary | |
| 20 | Muse Spark 1.1 | Meta | 1 489 | 1 480–1 498 | 4 579 | Proprietary |
| 21 | Gemini 3.6 Flash в режиме high | 1 487 | 1 478–1 496 | 4 226 | Proprietary | |
| 22 | Gemini 3 Pro | 1 485 | 1 477–1 493 | 6 682 | Proprietary | |
| 23 | GLM 5.3 Flash | Z.ai | 1 485 | 1 470–1 500 | 1 580 | MIT |
| 24 | GPT-5.5 в режиме high | OpenAI | 1 484 | 1 478–1 491 | 10 840 | Proprietary |
| 25 | ERNIE 5.1 | Baidu | 1 484 | 1 476–1 493 | 5 922 | Proprietary |
| 26 | GPT-5.5 · 5.5 | OpenAI | 1 483 | 1 476–1 490 | 10 943 | Proprietary |
| 27 | MiMo-V2.5-Pro | Xiaomi | 1 483 | 1 476–1 489 | 9 903 | MIT |
| 28 | GLM 5.2 в режиме max | Z.ai | 1 482 | 1 473–1 490 | 5 983 | MIT |
| 29 | Gemini 2.5 Pro | 1 480 | 1 476–1 485 | 20 322 | Proprietary | |
| 30 | Qwen3.7 Max | Alibaba Qwen | 1 480 | 1 456–1 503 | 654 | Proprietary |
| 31 | GLM 5.1 | Z.ai | 1 478 | 1 471–1 485 | 8 005 | MIT |
| 32 | GPT-5.4 в режиме high | OpenAI | 1 478 | 1 471–1 485 | 9 869 | Proprietary |
| 33 | Muse Spark | Meta | 1 478 | 1 465–1 490 | 2 264 | Proprietary |
| 34 | Claude Sonnet 4.6 | Anthropic | 1 476 | 1 469–1 482 | 10 779 | Proprietary |
| 35 | Claude Opus 4.8 в режиме high | Anthropic | 1 475 | 1 468–1 482 | 8 466 | Proprietary |
| 36 | Gemini 3 Flash Preview | 1 473 | 1 464–1 482 | 4 863 | Proprietary | |
| 37 | Qwen3.8 27B | Alibaba Qwen | 1 472 | 1 458–1 487 | 1 647 | Apache 2.0 |
| 38 | DeepSeek V4 Pro 0423 | DeepSeek | 1 469 | 1 462–1 477 | 8 597 | MIT |
| 39 | Nemotron 3 Ultra | NVIDIA | 1 469 | 1 455–1 483 | 1 745 | OpenMDW-1.1 |
| 40 | Claude Opus 4.8 | Anthropic | 1 469 | 1 462–1 476 | 8 757 | Proprietary |
| 41 | Kimi K2.6 | Moonshot AI | 1 468 | 1 459–1 476 | 6 226 | Modified MIT |
| 42 | Grok 4.5 | xAI | 1 467 | 1 458–1 476 | 4 865 | Proprietary |
| 43 | ERNIE 5.0 · preview-1203 | Baidu | 1 466 | 1 452–1 481 | 1 599 | Proprietary |
| 44 | Qwen3.6 Max Preview | Alibaba Qwen | 1 466 | 1 446–1 486 | 880 | Proprietary |
| 45 | Qwen3.7 Plus | Alibaba Qwen | 1 464 | 1 456–1 472 | 6 254 | Proprietary |
| 46 | GPT-5.4 | OpenAI | 1 463 | 1 456–1 470 | 10 290 | Proprietary |
| 47 | Hy3 | Tencent | 1 463 | 1 446–1 480 | 1 223 | Apache 2.0 |
| 48 | Claude Sonnet 5 в режиме high | Anthropic | 1 462 | 1 453–1 470 | 5 793 | Proprietary |
| 49 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 462 | 1 454–1 469 | 8 144 | MIT |
| 50 | GLM 5 | Z.ai | 1 460 | 1 451–1 469 | 4 422 | MIT |
| 51 | Grok 4.20 · beta-0309-reasoning | xAI | 1 460 | 1 454–1 467 | 10 081 | Proprietary |
| 52 | Kimi K2.5 · thinking | Moonshot AI | 1 460 | 1 454–1 466 | 11 016 | Modified MIT |
| 53 | GPT-6 Astra в режиме max | OpenAI | 1 459 | 1 431–1 487 | 427 | Proprietary |
| 54 | GPT-5.1 в режиме high | OpenAI | 1 458 | 1 450–1 466 | 6 628 | Proprietary |
| 55 | Qwen3 Max · preview | Alibaba Qwen | 1 458 | 1 448–1 467 | 4 192 | Proprietary |
| 56 | MiMo-V2 Pro | Xiaomi | 1 457 | 1 447–1 467 | 3 913 | Proprietary |
| 57 | GLM 4.6 | Z.ai | 1 457 | 1 449–1 465 | 5 810 | MIT |
| 58 | Qwen3.5 397B A17B | Alibaba Qwen | 1 457 | 1 451–1 463 | 12 478 | Apache 2.0 |
| 59 | GLM 4.7 | Z.ai | 1 456 | 1 443–1 469 | 2 064 | MIT |
| 60 | Grok 4.20 Multi-Agent | xAI | 1 455 | 1 448–1 462 | 9 903 | Proprietary |
| 61 | Gemini 3.5 Flash Lite | 1 455 | 1 445–1 464 | 4 262 | Proprietary | |
| 62 | Inkling | Thinking Machines Lab | 1 454 | 1 445–1 464 | 4 157 | Apache 2.0 |
| 63 | Gemma 4 31B | 1 454 | 1 435–1 474 | 884 | Apache 2.0 | |
| 64 | Claude Opus 4.5 | Anthropic | 1 454 | 1 447–1 460 | 11 020 | Proprietary |
| 65 | Gemini 3 Flash Preview в режиме minimal | 1 454 | 1 448–1 459 | 13 943 | Proprietary | |
| 66 | MiniMax M3 | MiniMax | 1 453 | 1 446–1 461 | 7 689 | MiniMax Community License |
| 67 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 453 | 1 443–1 462 | 4 570 | Proprietary |
| 68 | Seed 2.0 Pro | ByteDance Seed | 1 453 | 1 446–1 459 | 11 856 | Proprietary |
| 69 | DeepSeek V3.1 Terminus · terminus | DeepSeek | 1 452 | 1 428–1 476 | 572 | MIT |
| 70 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 452 | 1 443–1 462 | 4 331 | Proprietary |
| 71 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 451 | 1 435–1 467 | 1 394 | MIT |
| 72 | ERNIE 5.0 · 0110 | Baidu | 1 451 | 1 443–1 458 | 5 808 | Proprietary |
| 73 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 450 | 1 426–1 474 | 516 | Proprietary |
| 74 | DeepSeek V3.1 Terminus · terminus-thinking | DeepSeek | 1 450 | 1 425–1 475 | 526 | MIT |
| 75 | ERNIE 5.0 · preview-1022 | Baidu | 1 449 | 1 429–1 470 | 772 | Proprietary |
| 76 | GLM 5V Turbo | Z.ai | 1 449 | 1 433–1 464 | 1 459 | Proprietary |
| 77 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 449 | 1 426–1 472 | 597 | Proprietary |
| 78 | DeepSeek V4 Flash 0423 | DeepSeek | 1 449 | 1 441–1 456 | 7 778 | MIT |
| 79 | Claude Sonnet 4.5 | Anthropic | 1 447 | 1 442–1 453 | 13 001 | Proprietary |
| 80 | Mistral Medium 3.1 | Mistral AI | 1 446 | 1 441–1 451 | 14 806 | Proprietary |
| 81 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 446 | 1 430–1 461 | 1 418 | MIT |
| 82 | Claude Opus 4.5 в режиме high | Anthropic | 1 446 | 1 437–1 454 | 5 786 | Proprietary |
| 83 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 445 | 1 438–1 452 | 7 318 | MIT |
| 84 | Mistral Large 3 2512 | Mistral AI | 1 445 | 1 439–1 451 | 11 175 | Apache 2.0 |
| 85 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 445 | 1 435–1 454 | 3 887 | Proprietary |
| 86 | Grok 4.6 в режиме high | xAI | 1 444 | 1 432–1 456 | 2 495 | Proprietary |
| 87 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 444 | 1 429–1 459 | 1 724 | Apache 2.0 |
| 88 | Gemma 4 26B A4B | 1 444 | 1 425–1 463 | 876 | Apache 2.0 | |
| 89 | Grok 4.1 · 4.1 | xAI | 1 444 | 1 437–1 450 | 10 643 | Proprietary |
| 90 | GLM 4.5 | Z.ai | 1 443 | 1 434–1 453 | 3 855 | MIT |
| 91 | ChatGPT-4o (latest) | OpenAI | 1 442 | 1 437–1 448 | 13 506 | Proprietary |
| 92 | Grok 4.20 · beta1 | xAI | 1 442 | 1 433–1 451 | 4 373 | Proprietary |
| 93 | Qwen3.6 Plus | Alibaba Qwen | 1 442 | 1 434–1 449 | 7 137 | Proprietary |
| 94 | MiMo-V2.5 | Xiaomi | 1 440 | 1 432–1 448 | 7 193 | MIT |
| 95 | GPT-5.2 Chat | OpenAI | 1 440 | 1 432–1 448 | 5 467 | Proprietary |
| 96 | GPT-5.1 | OpenAI | 1 439 | 1 432–1 447 | 6 942 | Proprietary |
| 97 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 439 | 1 434–1 445 | 13 053 | Proprietary |
| 98 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 439 | 1 434–1 444 | 15 263 | Apache 2.0 |
| 99 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 439 | 1 431–1 446 | 7 951 | MIT |
| 100 | Grok 4.1 · 4.1-thinking | xAI | 1 438 | 1 432–1 445 | 10 665 | Proprietary |
| 101 | Grok 3 | xAI | 1 437 | 1 429–1 445 | 5 715 | Proprietary |
| 102 | Qwen3.5-27B | Alibaba Qwen | 1 435 | 1 426–1 444 | 4 500 | Apache 2.0 |
| 103 | Grok 4 | xAI | 1 435 | 1 427–1 443 | 6 644 | Proprietary |
| 104 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 435 | 1 426–1 443 | 5 117 | Proprietary | |
| 105 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 434 | 1 424–1 444 | 3 503 | Apache 2.0 |
| 106 | Gemini 2.5 Flash · flash | 1 434 | 1 429–1 438 | 20 029 | Proprietary | |
| 107 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 434 | 1 418–1 449 | 1 490 | Apache 2.0 |
| 108 | Hunyuan T1 | Tencent | 1 433 | 1 411–1 455 | 722 | Proprietary |
| 109 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 433 | 1 421–1 445 | 2 289 | MIT |
| 110 | MiMo-V2 Omni | Xiaomi | 1 433 | 1 422–1 444 | 3 049 | Proprietary |
| 111 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 432 | 1 424–1 441 | 4 638 | Apache 2.0 |
| 112 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 432 | 1 423–1 441 | 4 573 | Proprietary |
| 113 | Inkling Small | Thinking Machines Lab | 1 431 | 1 420–1 442 | 3 146 | Apache 2.0 |
| 114 | R1 0528 | DeepSeek | 1 431 | 1 420–1 441 | 3 378 | MIT |
| 115 | Kimi K2.5 · instant | Moonshot AI | 1 430 | 1 414–1 447 | 1 186 | Modified MIT |
| 116 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 429 | 1 420–1 438 | 4 592 | Proprietary |
| 117 | o3 | OpenAI | 1 429 | 1 422–1 435 | 9 898 | Proprietary |
| 118 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 428 | 1 414–1 441 | 1 914 | MIT |
| 119 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 428 | 1 420–1 435 | 6 570 | MIT |
| 120 | GPT-5.5 · 5.5-instant | OpenAI | 1 428 | 1 418–1 437 | 4 234 | Proprietary |
| 121 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 427 | 1 413–1 440 | 1 834 | MIT |
| 122 | Gemini 3.1 Flash Lite Preview | 1 427 | 1 420–1 433 | 9 874 | Proprietary | |
| 123 | Grok 4 Fast · reasoning | xAI | 1 426 | 1 415–1 437 | 2 917 | Proprietary |
| 124 | LongCat-Flash-Chat · chat | Meituan | 1 426 | 1 412–1 440 | 1 681 | MIT |
| 125 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 426 | 1 412–1 439 | 1 850 | Proprietary |
| 126 | Mistral Medium 3.5 | Mistral AI | 1 425 | 1 411–1 440 | 1 818 | Modified MIT |
| 127 | GPT-5.2 в режиме high | OpenAI | 1 424 | 1 417–1 431 | 7 829 | Proprietary |
| 128 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 423 | 1 416–1 430 | 7 473 | MIT |
| 129 | GPT-5.2 | OpenAI | 1 421 | 1 415–1 428 | 12 912 | Proprietary |
| 130 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 421 | 1 414–1 428 | 7 824 | Proprietary |
| 131 | Hy3 preview | Tencent | 1 421 | 1 403–1 438 | 1 202 | tencent-hunyuan-community |
| 132 | GPT-4.5 Preview | OpenAI | 1 420 | 1 408–1 432 | 2 545 | Proprietary |
| 133 | Grok 4.1 Fast | xAI | 1 419 | 1 412–1 425 | 9 039 | Proprietary |
| 134 | Kimi K2 Thinking | Moonshot AI | 1 418 | 1 412–1 425 | 9 860 | Modified MIT |
| 135 | GPT-5.4 Mini в режиме high | OpenAI | 1 418 | 1 411–1 425 | 9 773 | Proprietary |
| 136 | GPT-5 | OpenAI | 1 417 | 1 408–1 425 | 4 985 | Proprietary |
| 137 | Step 3.5 Flash | StepFun | 1 417 | 1 410–1 423 | 9 280 | Apache 2.0 |
| 138 | Qwen3.5-Flash | Alibaba Qwen | 1 416 | 1 409–1 423 | 9 205 | Proprietary |
| 139 | Claude Opus 4.1 · 20250805 | Anthropic | 1 416 | 1 410–1 422 | 12 040 | Proprietary |
| 140 | MiniMax M2.7 | MiniMax | 1 416 | 1 409–1 422 | 10 908 | Modified MIT |
| 141 | MiniMax M2.1 | MiniMax | 1 415 | 1 404–1 427 | 2 629 | MIT |
| 142 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 415 | 1 400–1 431 | 1 414 | Proprietary |
| 143 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 414 | 1 397–1 430 | 1 224 | Apache 2.0 |
| 144 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 412 | 1 403–1 421 | 4 937 | Apache 2.0 |
| 145 | Hunyuan Vision 1.5 | Tencent | 1 410 | 1 380–1 441 | 337 | Proprietary |
| 146 | Grok 4 Fast · chat | xAI | 1 407 | 1 388–1 427 | 957 | Proprietary |
| 147 | GLM 4.5 Air | Z.ai | 1 407 | 1 399–1 416 | 4 875 | MIT |
| 148 | Amazon Nova Experimental Chat · 10-09 | Amazon | 1 406 | 1 379–1 434 | 428 | Proprietary |
| 149 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 406 | 1 398–1 414 | 6 233 | Apache 2.0 |
| 150 | Claude Haiku 4.5 | Anthropic | 1 404 | 1 400–1 409 | 20 863 | Proprietary |
| 151 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 404 | 1 381–1 428 | 554 | Proprietary |
| 152 | GPT-5 в режиме high | OpenAI | 1 403 | 1 394–1 411 | 5 085 | Proprietary |
| 153 | Grok 4.3 | xAI | 1 401 | 1 394–1 408 | 10 976 | Proprietary |
| 154 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 401 | 1 386–1 415 | 1 678 | MIT |
| 155 | Hunyuan TurboS · 20250416 | Tencent | 1 400 | 1 387–1 414 | 1 944 | Proprietary |
| 156 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 396 | 1 389–1 403 | 7 394 | Proprietary | |
| 157 | GLM 4.6V | Z.ai | 1 394 | 1 367–1 421 | 462 | MIT |
| 158 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 393 | 1 383–1 402 | 3 653 | Apache 2.0 |
| 159 | Muse Glimmer 30B | Meta | 1 392 | 1 368–1 416 | 625 | Apache-2.0 |
| 160 | Nemotron 3 Super | NVIDIA | 1 388 | 1 371–1 405 | 1 136 | NVIDIA Open Model |
| 161 | Solar Pro 4 | Upstage | 1 387 | 1 357–1 417 | 394 | Proprietary |
| 162 | Granite 4.2 30B | IBM Granite | 1 387 | 1 360–1 414 | 485 | Apache 2.0 |
| 163 | Kimi K2 0905 | Moonshot AI | 1 386 | 1 372–1 400 | 1 737 | Modified MIT |
| 164 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 384 | 1 371–1 397 | 2 095 | Apache 2.0 |
| 165 | Gemma 3 27B | 1 383 | 1 376–1 390 | 7 849 | Gemma | |
| 166 | GPT-5.3 Chat | OpenAI | 1 383 | 1 374–1 392 | 5 169 | Proprietary |
| 167 | Nova 2 Lite | Amazon | 1 383 | 1 370–1 396 | 2 062 | Proprietary |
| 168 | GPT-4.1 | OpenAI | 1 382 | 1 375–1 389 | 8 439 | Proprietary |
| 169 | DeepSeek V3 0324 | DeepSeek | 1 382 | 1 375–1 389 | 7 611 | MIT |
| 170 | Qwen2.5 Max | Alibaba Qwen | 1 381 | 1 373–1 390 | 5 665 | Proprietary |
| 171 | gpt-oss-120b | OpenAI | 1 381 | 1 372–1 390 | 4 791 | Apache 2.0 |
| 172 | Grok 3 Mini в режиме high | xAI | 1 381 | 1 369–1 392 | 2 577 | Proprietary |
| 173 | Mistral Medium 3 | Mistral AI | 1 381 | 1 372–1 389 | 5 615 | Proprietary |
| 174 | Ling-flash-2.0 | inclusionAI | 1 379 | 1 360–1 397 | 995 | MIT |
| 175 | GPT-5.4 Nano в режиме high | OpenAI | 1 377 | 1 369–1 384 | 9 782 | Proprietary |
| 176 | R1 | DeepSeek | 1 375 | 1 365–1 385 | 3 309 | MIT |
| 177 | Grok 3 Mini | xAI | 1 373 | 1 363–1 383 | 3 647 | Proprietary |
| 178 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 373 | 1 365–1 382 | 5 191 | Proprietary | |
| 179 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 372 | 1 363–1 381 | 4 550 | Apache 2.0 |
| 180 | Qwen3 32B | Alibaba Qwen | 1 372 | 1 352–1 392 | 798 | Apache 2.0 |
| 181 | GPT-5 Mini в режиме high | OpenAI | 1 371 | 1 362–1 380 | 4 179 | Proprietary |
| 182 | MiniMax M2.5 | MiniMax | 1 369 | 1 361–1 377 | 6 739 | Modified MIT |
| 183 | o1 · 2024-12-17 | OpenAI | 1 369 | 1 360–1 378 | 4 820 | Proprietary |
| 184 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 367 | 1 344–1 390 | 588 | Nvidia Open |
| 185 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 366 | 1 358–1 375 | 5 896 | Proprietary |
| 186 | Gemma 3 12B | 1 366 | 1 345–1 386 | 760 | Gemma | |
| 187 | Kimi K2 0711 | Moonshot AI | 1 364 | 1 355–1 373 | 4 340 | Modified MIT |
| 188 | Claude Opus 4 · 20250514 | Anthropic | 1 362 | 1 354–1 370 | 6 959 | Proprietary |
| 189 | INTELLECT-3 | Prime Intellect | 1 361 | 1 341–1 381 | 884 | MIT |
| 190 | Gemini 2.0 Flash | 1 361 | 1 353–1 368 | 7 776 | Proprietary | |
| 191 | GLM 4.7 Flash | Z.ai | 1 360 | 1 347–1 374 | 1 773 | MIT |
| 192 | Step-1o Turbo | StepFun | 1 358 | 1 343–1 373 | 1 500 | Proprietary |
| 193 | Trinity Large Thinking | Arcee AI | 1 358 | 1 348–1 367 | 4 838 | Apache 2.0 |
| 194 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 355 | 1 343–1 367 | 2 467 | NVIDIA Open Model |
| 195 | MiniMax M1 | MiniMax | 1 354 | 1 346–1 362 | 5 767 | Apache 2.0 |
| 196 | o4 Mini | OpenAI | 1 354 | 1 346–1 361 | 7 510 | Proprietary |
| 197 | Step 3 | StepFun | 1 353 | 1 335–1 372 | 1 065 | Apache 2.0 |
| 198 | Trinity Large | Arcee AI | 1 353 | 1 344–1 362 | 4 869 | Apache 2.0 |
| 199 | Qwen-Plus | Alibaba Qwen | 1 352 | 1 334–1 371 | 968 | Proprietary |
| 200 | GLM-4-Plus · plus-0111 | Z.ai | 1 350 | 1 332–1 369 | 986 | Proprietary |
| 201 | Mercury 2 | Inception Labs | 1 350 | 1 324–1 377 | 461 | Proprietary |
| 202 | MiniMax M2 | MiniMax | 1 350 | 1 332–1 368 | 1 103 | Apache 2.0 |
| 203 | o1 · preview | OpenAI | 1 348 | 1 339–1 357 | 5 419 | Proprietary |
| 204 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 348 | 1 338–1 358 | 4 014 | Apache 2.0 |
| 205 | QwQ 32B | Alibaba Qwen | 1 347 | 1 339–1 356 | 4 509 | Apache 2.0 |
| 206 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 345 | 1 337–1 353 | 5 596 | Proprietary |
| 207 | Mistral Small 3.2 24B | Mistral AI | 1 340 | 1 329–1 351 | 2 883 | Apache 2.0 |
| 208 | Claude Sonnet 4 · 20250514 | Anthropic | 1 340 | 1 332–1 348 | 6 347 | Proprietary |
| 209 | o3 Mini High | OpenAI | 1 339 | 1 329–1 350 | 3 382 | Proprietary |
| 210 | Command A | Cohere | 1 339 | 1 332–1 345 | 9 338 | CC-BY-NC-4.0 |
| 211 | Ring-flash-2.0 | inclusionAI | 1 337 | 1 319–1 355 | 1 072 | MIT |
| 212 | Granite 4.2 8B | IBM Granite | 1 334 | 1 307–1 361 | 502 | Apache 2.0 |
| 213 | DeepSeek V3 | DeepSeek | 1 333 | 1 323–1 342 | 3 766 | DeepSeek |
| 214 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 332 | 1 305–1 360 | 395 | Nvidia |
| 215 | Gemini 2.0 Flash-Lite | 1 332 | 1 323–1 341 | 4 444 | Proprietary | |
| 216 | Step-2 16k | StepFun | 1 330 | 1 310–1 349 | 863 | Proprietary |
| 217 | GLM 4.5V | Z.ai | 1 330 | 1 308–1 351 | 738 | MIT |
| 218 | GPT-4.1 Mini | OpenAI | 1 328 | 1 320–1 336 | 6 481 | Proprietary |
| 219 | Hunyuan Turbo | Tencent | 1 328 | 1 301–1 355 | 414 | Proprietary |
| 220 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 1 324 | 1 298–1 350 | 498 | Nvidia Open Model |
| 221 | Gemma 3n E4B | 1 324 | 1 314–1 334 | 3 668 | Gemma | |
| 222 | Qwen3 30B A3B | Alibaba Qwen | 1 321 | 1 312–1 330 | 4 579 | Apache 2.0 |
| 223 | Gemini 1.5 Pro · 1.5-pro-002 | 1 321 | 1 314–1 327 | 9 790 | Proprietary | |
| 224 | Nemotron 3.5 Lightning | NVIDIA | 1 320 | 1 302–1 337 | 1 342 | OpenMDW-1.1 |
| 225 | GPT-5 Nano в режиме high | OpenAI | 1 318 | 1 302–1 335 | 1 314 | Proprietary |
| 226 | Hunyuan TurboS · 20250226 | Tencent | 1 318 | 1 290–1 345 | 384 | Proprietary |
| 227 | Qwen2.5 Plus | Alibaba Qwen | 1 317 | 1 304–1 331 | 1 771 | Proprietary |
| 228 | o1-mini | OpenAI | 1 316 | 1 308–1 323 | 8 979 | Proprietary |
| 229 | Yi-Lightning | 01.AI | 1 315 | 1 306–1 325 | 4 822 | Proprietary |
| 230 | o3 Mini | OpenAI | 1 315 | 1 309–1 322 | 9 602 | Proprietary |
| 231 | OLMo 3.1 32B Instruct | Ai2 | 1 315 | 1 300–1 329 | 1 795 | Apache 2.0 |
| 232 | OLMo 3 32B Think | Ai2 | 1 314 | 1 295–1 334 | 982 | Apache 2.0 |
| 233 | Grok 2 | xAI | 1 314 | 1 307–1 321 | 10 796 | Proprietary |
| 234 | Granite 4.2 3B | IBM Granite | 1 313 | 1 283–1 343 | 467 | Apache 2.0 |
| 235 | Athene V2 Chat | Nexusflow | 1 307 | 1 298–1 316 | 4 368 | NexusFlow |
| 236 | Gemma 3 4B | 1 306 | 1 285–1 326 | 780 | Gemma | |
| 237 | Granite 4.1 8B | IBM Granite | 1 303 | 1 278–1 328 | 666 | Apache 2.0 |
| 238 | Grok 2 Mini | xAI | 1 300 | 1 292–1 307 | 8 788 | Proprietary |
| 239 | Gemini 1.5 Flash · 002 | 1 299 | 1 291–1 308 | 6 083 | Proprietary | |
| 240 | Mercury | Inception Labs | 1 299 | 1 265–1 333 | 294 | Proprietary |
| 241 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 299 | 1 291–1 307 | 6 691 | Proprietary |
| 242 | GPT-4.1 Nano | OpenAI | 1 297 | 1 281–1 314 | 1 162 | Proprietary |
| 243 | gpt-oss-20b | OpenAI | 1 296 | 1 281–1 310 | 1 669 | Apache 2.0 |
| 244 | Llama 3.1 405B Instruct · bf16 | Meta | 1 295 | 1 288–1 302 | 7 177 | Llama 3.1 Community |
| 245 | GPT-4o (2024-05-13) | OpenAI | 1 295 | 1 288–1 302 | 19 017 | Proprietary |
| 246 | GLM-4-Plus · plus | Z.ai | 1 295 | 1 285–1 304 | 4 702 | Proprietary |
| 247 | Athene 70B | Nexusflow | 1 294 | 1 282–1 305 | 3 029 | CC-BY-NC-4.0 |
| 248 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 293 | 1 277–1 309 | 1 299 | Llama 3.1 |
| 249 | Hunyuan Large | Tencent | 1 291 | 1 270–1 312 | 707 | Proprietary |
| 250 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 288 | 1 271–1 305 | 1 137 | DeepSeek |
| 251 | OLMo 3.1 32B Think | Ai2 | 1 287 | 1 271–1 304 | 1 367 | Apache 2.0 |
| 252 | Llama 4 Maverick | Meta | 1 287 | 1 279–1 295 | 6 703 | Llama 4 |
| 253 | GPT-4o-mini (2024-07-18) | OpenAI | 1 286 | 1 279–1 293 | 11 641 | Proprietary |
| 254 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 285 | 1 277–1 293 | 7 420 | Proprietary |
| 255 | Hunyuan Standard | Tencent | 1 284 | 1 263–1 306 | 696 | Proprietary |
| 256 | Llama 3.1 405B Instruct · fp8 | Meta | 1 284 | 1 277–1 292 | 10 001 | Llama 3.1 Community |
| 257 | Llama 3.3 70B Instruct | Meta | 1 284 | 1 277–1 290 | 9 340 | Llama-3.3 |
| 258 | Qwen Max | Alibaba Qwen | 1 279 | 1 268–1 290 | 2 975 | Qwen |
| 259 | Llama 4 Scout | Meta | 1 279 | 1 270–1 288 | 4 968 | Llama |
| 260 | Llama 3.1 70B Instruct | Meta | 1 276 | 1 269–1 283 | 9 381 | Llama 3.1 Community |
| 261 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 275 | 1 267–1 283 | 6 932 | Qwen |
| 262 | GPT-4o (2024-08-06) | OpenAI | 1 274 | 1 266–1 282 | 7 560 | Proprietary |
| 263 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 273 | 1 264–1 283 | 4 318 | DeepSeek |
| 264 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 270 | 1 264–1 276 | 15 548 | Proprietary |
| 265 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 270 | 1 263–1 277 | 13 749 | Proprietary |
| 266 | Gemini 1.5 Pro · advanced-0514 | 1 268 | 1 258–1 277 | 8 815 | Proprietary | |
| 267 | Gemini 1.5 Pro · 1.5-pro-001 | 1 267 | 1 259–1 275 | 13 570 | Proprietary | |
| 268 | Llama 3.1 Tulu 3 70B | Ai2 | 1 267 | 1 243–1 291 | 520 | Llama 3.1 |
| 269 | Mistral Large 2407 | Mistral AI | 1 266 | 1 258–1 274 | 7 485 | Mistral Research |
| 270 | Mistral Small 3.1 24B | Mistral AI | 1 266 | 1 257–1 274 | 5 210 | Apache 2.0 |
| 271 | Reka Core | Reka AI | 1 265 | 1 249–1 282 | 1 183 | Proprietary |
| 272 | Mistral Large | Mistral AI | 1 260 | 1 252–1 269 | 4 869 | MRL |
| 273 | Hunyuan Large Vision | Tencent | 1 259 | 1 238–1 279 | 911 | Proprietary |
| 274 | GPT-4 Turbo | OpenAI | 1 256 | 1 248–1 263 | 16 354 | Proprietary |
| 275 | Nova Pro 1.0 | Amazon | 1 255 | 1 246–1 264 | 4 281 | Proprietary |
| 276 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 255 | 1 247–1 262 | 17 680 | Proprietary |
| 277 | Claude 3 Opus | Anthropic | 1 254 | 1 248–1 260 | 33 790 | Proprietary |
| 278 | Granite 4.0 H Small | IBM Granite | 1 252 | 1 231–1 274 | 877 | Apache 2.0 |
| 279 | Command R+ (08-2024) | Cohere | 1 252 | 1 237–1 266 | 1 639 | CC-BY-NC-4.0 |
| 280 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 250 | 1 242–1 258 | 16 044 | Proprietary |
| 281 | Jamba 1.5 Large | AI21 Labs | 1 249 | 1 233–1 265 | 1 358 | Jamba Open |
| 282 | Claude 3.5 Haiku | Anthropic | 1 247 | 1 241–1 253 | 11 835 | Proprietary |
| 283 | OLMo 2 32B Instruct | Ai2 | 1 240 | 1 216–1 263 | 580 | Apache-2.0 |
| 284 | Mistral Small 3 | Mistral AI | 1 235 | 1 223–1 247 | 2 579 | Apache 2.0 |
| 285 | Magistral Medium | Mistral AI | 1 234 | 1 219–1 249 | 1 926 | Proprietary |
| 286 | Aya Expanse 32B | Cohere | 1 232 | 1 223–1 242 | 4 768 | CC-BY-NC-4.0 |
| 287 | Nova Lite 1.0 | Amazon | 1 232 | 1 222–1 243 | 3 333 | Proprietary |
| 288 | Gemini 1.5 Flash-8B | 1 232 | 1 224–1 240 | 6 236 | Proprietary | |
| 289 | Gemini 1.5 Flash · 001 | 1 232 | 1 224–1 240 | 10 875 | Proprietary | |
| 290 | Gemma 2 9B IT SimPO | Princeton NLP | 1 231 | 1 216–1 246 | 1 596 | MIT |
| 291 | Nemotron-4 340B Instruct | NVIDIA | 1 228 | 1 217–1 240 | 3 316 | NVIDIA Open Model |
| 292 | Gemma 2 27B | 1 228 | 1 221–1 235 | 12 863 | Gemma license | |
| 293 | GLM-4 | Z.ai | 1 226 | 1 211–1 241 | 1 590 | Proprietary |
| 294 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 226 | 1 204–1 247 | 682 | Llama 3.1 |
| 295 | Llama 3 70B Instruct | Meta | 1 226 | 1 218–1 233 | 25 903 | Llama 3 Community |
| 296 | Reka Flash (2024-09) | Reka AI | 1 223 | 1 207–1 239 | 1 277 | Proprietary |
| 297 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 221 | 1 203–1 239 | 956 | Apache 2.0 |
| 298 | Command R (08-2024) | Cohere | 1 218 | 1 204–1 233 | 1 620 | CC-BY-NC-4.0 |
| 299 | Command R+ | Cohere | 1 216 | 1 208–1 225 | 13 227 | CC-BY-NC-4.0 |
| 300 | Nova Micro 1.0 | Amazon | 1 210 | 1 200–1 221 | 3 314 | 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 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.