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 «длинные запросы», 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 | 1 523 | 1 511–1 536 | 2 413 | Proprietary |
| 2 | Claude Opus 4.6 в режиме high | Anthropic | 1 520 | 1 515–1 525 | 29 738 | Proprietary |
| 3 | Claude Opus 5 в режиме high | Anthropic | 1 517 | 1 511–1 523 | 20 232 | Proprietary |
| 4 | Claude Opus 4.6 | Anthropic | 1 516 | 1 511–1 521 | 32 179 | Proprietary |
| 5 | Claude Opus 5 в режиме max | Anthropic | 1 515 | 1 507–1 522 | 9 770 | Proprietary |
| 6 | Claude Fable 5 | Anthropic | 1 509 | 1 503–1 515 | 13 819 | Proprietary |
| 7 | Gemini 3.8 Flash в режиме high | 1 506 | 1 494–1 518 | 2 538 | Proprietary | |
| 8 | Claude Opus 4.7 в режиме high | Anthropic | 1 504 | 1 499–1 510 | 27 145 | Proprietary |
| 9 | Muse Spark 1.3 в режиме max | Meta | 1 500 | 1 486–1 513 | 2 151 | Proprietary |
| 10 | Claude Opus 4.7 | Anthropic | 1 499 | 1 493–1 504 | 27 866 | Proprietary |
| 11 | Gemini 3.7 Flash в режиме high | 1 492 | 1 480–1 504 | 2 680 | Proprietary | |
| 12 | Kimi K3 в режиме max | Moonshot AI | 1 491 | 1 484–1 498 | 9 244 | Kimi K3 license |
| 13 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 488 | 1 481–1 496 | 7 561 | Proprietary |
| 14 | GLM 5.3 в режиме max | Z.ai | 1 488 | 1 479–1 497 | 5 257 | MIT |
| 15 | Muse Spark 1.2 в режиме xhigh | Meta | 1 487 | 1 472–1 503 | 1 487 | Proprietary |
| 16 | Qwen3.7 Max | Alibaba Qwen | 1 484 | 1 469–1 500 | 1 607 | Proprietary |
| 17 | GPT-5.5 в режиме high | OpenAI | 1 484 | 1 479–1 490 | 29 404 | Proprietary |
| 18 | Gemini 3.1 Pro Preview | 1 483 | 1 479–1 488 | 45 996 | Proprietary | |
| 19 | MiMo-V2.5-Pro | Xiaomi | 1 482 | 1 477–1 487 | 26 794 | MIT |
| 20 | Claude Opus 4.8 в режиме high | Anthropic | 1 482 | 1 476–1 487 | 24 417 | Proprietary |
| 21 | Gemini 3.5 Flash в режиме high | 1 482 | 1 476–1 487 | 17 719 | Proprietary | |
| 22 | Claude Opus 4.8 | Anthropic | 1 481 | 1 475–1 486 | 24 899 | Proprietary |
| 23 | Claude Sonnet 4.6 | Anthropic | 1 480 | 1 475–1 486 | 28 350 | Proprietary |
| 24 | GPT-5.5 · 5.5 | OpenAI | 1 480 | 1 475–1 485 | 30 469 | Proprietary |
| 25 | Claude Opus 4.5 | Anthropic | 1 480 | 1 475–1 485 | 22 794 | Proprietary |
| 26 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 479 | 1 473–1 486 | 12 438 | Proprietary |
| 27 | GLM 5.3 Flash | Z.ai | 1 478 | 1 468–1 487 | 4 663 | MIT |
| 28 | Claude Opus 4.5 в режиме high | Anthropic | 1 477 | 1 471–1 484 | 9 303 | Proprietary |
| 29 | Qwen3.5 Max | Alibaba Qwen | 1 477 | 1 470–1 485 | 8 386 | Proprietary |
| 30 | Gemini 3.6 Flash в режиме high | 1 477 | 1 470–1 483 | 12 561 | Proprietary | |
| 31 | Claude Sonnet 4.5 | Anthropic | 1 476 | 1 472–1 481 | 25 022 | Proprietary |
| 32 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 475 | 1 471–1 480 | 24 656 | Proprietary |
| 33 | GPT-5.4 в режиме high | OpenAI | 1 475 | 1 469–1 480 | 25 968 | Proprietary |
| 34 | Gemini 3.5 Flash в режиме medium | 1 475 | 1 469–1 481 | 17 436 | Proprietary | |
| 35 | GLM 5.2 в режиме max | Z.ai | 1 475 | 1 469–1 481 | 16 836 | MIT |
| 36 | Gemini 3 Pro | 1 472 | 1 465–1 478 | 10 571 | Proprietary | |
| 37 | GLM 5.1 | Z.ai | 1 467 | 1 462–1 472 | 21 963 | MIT |
| 38 | Kimi K2.6 | Moonshot AI | 1 466 | 1 460–1 472 | 15 529 | Modified MIT |
| 39 | Grok 4.5 | xAI | 1 465 | 1 458–1 471 | 14 426 | Proprietary |
| 40 | GPT-5.4 | OpenAI | 1 465 | 1 459–1 470 | 27 343 | Proprietary |
| 41 | Claude Sonnet 5 в режиме high | Anthropic | 1 464 | 1 458–1 470 | 16 276 | Proprietary |
| 42 | ERNIE 5.1 | Baidu | 1 463 | 1 457–1 469 | 15 363 | Proprietary |
| 43 | Muse Spark 1.1 | Meta | 1 462 | 1 456–1 469 | 12 747 | Proprietary |
| 44 | Qwen3.6 Max Preview | Alibaba Qwen | 1 458 | 1 444–1 471 | 2 147 | Proprietary |
| 45 | DeepSeek V4 Pro 0423 | DeepSeek | 1 457 | 1 452–1 463 | 23 898 | MIT |
| 46 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 456 | 1 447–1 466 | 4 136 | MIT |
| 47 | MiMo-V2 Pro | Xiaomi | 1 455 | 1 448–1 462 | 9 443 | Proprietary |
| 48 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 455 | 1 449–1 461 | 11 217 | Proprietary |
| 49 | Qwen3.7 Plus | Alibaba Qwen | 1 454 | 1 448–1 460 | 17 911 | Proprietary |
| 50 | Grok 4.6 в режиме high | xAI | 1 453 | 1 445–1 461 | 7 372 | Proprietary |
| 51 | GPT-6 Astra в режиме max | OpenAI | 1 453 | 1 435–1 471 | 1 177 | Proprietary |
| 52 | Gemini 3 Flash Preview | 1 452 | 1 445–1 460 | 7 966 | Proprietary | |
| 53 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 452 | 1 446–1 459 | 13 080 | Proprietary |
| 54 | Qwen3.8 27B | Alibaba Qwen | 1 450 | 1 441–1 459 | 4 876 | Apache 2.0 |
| 55 | Muse Spark | Meta | 1 449 | 1 440–1 458 | 5 308 | Proprietary |
| 56 | Gemini 2.5 Pro | 1 449 | 1 445–1 453 | 32 764 | Proprietary | |
| 57 | GPT-5.1 в режиме high | OpenAI | 1 447 | 1 441–1 454 | 10 373 | Proprietary |
| 58 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 447 | 1 441–1 453 | 22 755 | MIT |
| 59 | GLM 5 | Z.ai | 1 447 | 1 440–1 453 | 10 267 | MIT |
| 60 | MiniMax M3 | MiniMax | 1 446 | 1 441–1 452 | 21 742 | MiniMax Community License |
| 61 | Kimi K2.5 · thinking | Moonshot AI | 1 445 | 1 440–1 450 | 27 419 | Modified MIT |
| 62 | MiMo-V2.5 | Xiaomi | 1 445 | 1 439–1 451 | 19 657 | MIT |
| 63 | Claude Opus 4.1 · 20250805 | Anthropic | 1 445 | 1 439–1 450 | 17 952 | Proprietary |
| 64 | Gemma 4 31B | 1 444 | 1 430–1 457 | 1 698 | Apache 2.0 | |
| 65 | Hy3 | Tencent | 1 443 | 1 433–1 454 | 3 659 | Apache 2.0 |
| 66 | Qwen3.5 397B A17B | Alibaba Qwen | 1 443 | 1 438–1 448 | 33 206 | Apache 2.0 |
| 67 | MiMo-V2 Omni | Xiaomi | 1 440 | 1 432–1 448 | 8 513 | Proprietary |
| 68 | GLM 5V Turbo | Z.ai | 1 439 | 1 430–1 449 | 4 205 | Proprietary |
| 69 | Qwen3.6 Plus | Alibaba Qwen | 1 439 | 1 433–1 445 | 19 477 | Proprietary |
| 70 | Nemotron 3 Ultra | NVIDIA | 1 439 | 1 430–1 448 | 4 840 | OpenMDW-1.1 |
| 71 | Grok 3 | xAI | 1 439 | 1 430–1 447 | 4 862 | Proprietary |
| 72 | Qwen3 Max · preview | Alibaba Qwen | 1 439 | 1 431–1 446 | 6 101 | Proprietary |
| 73 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 438 | 1 419–1 456 | 958 | Proprietary |
| 74 | Kimi K2.5 · instant | Moonshot AI | 1 436 | 1 424–1 449 | 2 223 | Modified MIT |
| 75 | Grok 4.20 · beta-0309-reasoning | xAI | 1 436 | 1 431–1 441 | 26 619 | Proprietary |
| 76 | DeepSeek V4 Flash 0423 | DeepSeek | 1 435 | 1 429–1 441 | 21 502 | MIT |
| 77 | Gemini 3.5 Flash Lite | 1 435 | 1 428–1 441 | 12 487 | Proprietary | |
| 78 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 434 | 1 427–1 440 | 13 400 | Proprietary |
| 79 | GLM 4.7 | Z.ai | 1 431 | 1 421–1 442 | 3 112 | MIT |
| 80 | Inkling | Thinking Machines Lab | 1 431 | 1 424–1 438 | 12 221 | Apache 2.0 |
| 81 | GPT-5.1 | OpenAI | 1 431 | 1 424–1 437 | 11 610 | Proprietary |
| 82 | Grok 4.20 Multi-Agent | xAI | 1 429 | 1 424–1 435 | 25 589 | Proprietary |
| 83 | GPT-5.2 Chat | OpenAI | 1 429 | 1 423–1 435 | 13 016 | Proprietary |
| 84 | Gemma 4 26B A4B | 1 429 | 1 415–1 443 | 1 595 | Apache 2.0 | |
| 85 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 429 | 1 423–1 434 | 13 016 | MIT |
| 86 | Claude Haiku 4.5 | Anthropic | 1 428 | 1 424–1 432 | 47 202 | Proprietary |
| 87 | Gemini 3 Flash Preview в режиме minimal | 1 428 | 1 424–1 433 | 33 513 | Proprietary | |
| 88 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 427 | 1 423–1 431 | 26 046 | Apache 2.0 |
| 89 | Seed 2.0 Pro | ByteDance Seed | 1 427 | 1 422–1 432 | 29 956 | Proprietary |
| 90 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 426 | 1 420–1 432 | 11 115 | MIT |
| 91 | Grok 4.20 · beta1 | xAI | 1 425 | 1 419–1 432 | 9 851 | Proprietary |
| 92 | DeepSeek V3.1 Terminus · terminus-thinking | DeepSeek | 1 425 | 1 403–1 447 | 672 | MIT |
| 93 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 425 | 1 418–1 431 | 10 775 | Proprietary |
| 94 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 425 | 1 419–1 430 | 21 589 | MIT |
| 95 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 422 | 1 414–1 430 | 6 943 | Proprietary |
| 96 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 422 | 1 410–1 434 | 2 383 | MIT |
| 97 | GPT-5.5 · 5.5-instant | OpenAI | 1 422 | 1 415–1 429 | 11 768 | Proprietary |
| 98 | GLM 4.6 | Z.ai | 1 422 | 1 415–1 428 | 8 849 | MIT |
| 99 | ERNIE 5.0 · 0110 | Baidu | 1 421 | 1 415–1 427 | 11 174 | Proprietary |
| 100 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 421 | 1 408–1 434 | 2 224 | Apache 2.0 |
| 101 | Gemini 2.5 Flash · flash | 1 419 | 1 415–1 423 | 32 105 | Proprietary | |
| 102 | MiniMax M2.7 | MiniMax | 1 419 | 1 414–1 424 | 28 952 | Modified MIT |
| 103 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 417 | 1 406–1 427 | 2 978 | MIT |
| 104 | Mistral Medium 3.5 | Mistral AI | 1 416 | 1 407–1 425 | 4 644 | Modified MIT |
| 105 | Grok 4.1 · 4.1 | xAI | 1 415 | 1 410–1 421 | 20 068 | Proprietary |
| 106 | MiniMax M2.1 | MiniMax | 1 415 | 1 406–1 424 | 4 276 | MIT |
| 107 | Kimi K2 Thinking | Moonshot AI | 1 415 | 1 410–1 420 | 18 836 | Modified MIT |
| 108 | Grok 4 Fast · chat | xAI | 1 414 | 1 398–1 429 | 1 411 | Proprietary |
| 109 | GPT-5.2 | OpenAI | 1 413 | 1 409–1 418 | 31 246 | Proprietary |
| 110 | ChatGPT-4o (latest) | OpenAI | 1 413 | 1 408–1 418 | 17 302 | Proprietary |
| 111 | Mistral Large 3 2512 | Mistral AI | 1 413 | 1 408–1 417 | 24 893 | Apache 2.0 |
| 112 | Qwen3.5-27B | Alibaba Qwen | 1 412 | 1 406–1 419 | 10 344 | Apache 2.0 |
| 113 | GLM 4.5 | Z.ai | 1 412 | 1 404–1 421 | 5 068 | MIT |
| 114 | Grok 4 Fast · reasoning | xAI | 1 411 | 1 402–1 420 | 4 419 | Proprietary |
| 115 | ERNIE 5.0 · preview-1203 | Baidu | 1 411 | 1 399–1 423 | 2 464 | Proprietary |
| 116 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 410 | 1 404–1 417 | 11 006 | Apache 2.0 |
| 117 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 410 | 1 397–1 422 | 2 093 | MIT |
| 118 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 410 | 1 397–1 422 | 2 060 | Proprietary |
| 119 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 410 | 1 404–1 415 | 15 503 | MIT |
| 120 | GPT-5.2 в режиме high | OpenAI | 1 409 | 1 403–1 415 | 16 215 | Proprietary |
| 121 | Grok 4 | xAI | 1 409 | 1 402–1 415 | 8 774 | Proprietary |
| 122 | GPT-5.4 Mini в режиме high | OpenAI | 1 408 | 1 403–1 414 | 25 406 | Proprietary |
| 123 | ERNIE 5.0 · preview-1022 | Baidu | 1 408 | 1 391–1 424 | 1 168 | Proprietary |
| 124 | Hy3 preview | Tencent | 1 408 | 1 396–1 419 | 2 901 | tencent-hunyuan-community |
| 125 | Grok 4.1 · 4.1-thinking | xAI | 1 408 | 1 402–1 413 | 20 008 | Proprietary |
| 126 | Mistral Medium 3.1 | Mistral AI | 1 406 | 1 402–1 410 | 26 623 | Proprietary |
| 127 | GPT-4.5 Preview | OpenAI | 1 406 | 1 392–1 419 | 1 796 | Proprietary |
| 128 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 405 | 1 398–1 412 | 7 996 | Proprietary | |
| 129 | Hunyuan Vision 1.5 | Tencent | 1 404 | 1 380–1 429 | 513 | Proprietary |
| 130 | Step 3.5 Flash | StepFun | 1 404 | 1 398–1 409 | 21 164 | Apache 2.0 |
| 131 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 402 | 1 394–1 410 | 5 302 | Apache 2.0 |
| 132 | Claude Opus 4 · 20250514 | Anthropic | 1 401 | 1 394–1 409 | 8 345 | Proprietary |
| 133 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 401 | 1 382–1 420 | 888 | Proprietary |
| 134 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 400 | 1 389–1 410 | 3 246 | MIT |
| 135 | GPT-5 | OpenAI | 1 399 | 1 392–1 407 | 6 775 | Proprietary |
| 136 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 399 | 1 385–1 413 | 1 632 | Apache 2.0 |
| 137 | Inkling Small | Thinking Machines Lab | 1 398 | 1 391–1 405 | 9 108 | Apache 2.0 |
| 138 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 397 | 1 390–1 405 | 6 693 | Proprietary |
| 139 | DeepSeek V3.1 Terminus · terminus | DeepSeek | 1 396 | 1 376–1 415 | 843 | MIT |
| 140 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 395 | 1 388–1 403 | 6 169 | Proprietary |
| 141 | GPT-5.3 Chat | OpenAI | 1 394 | 1 388–1 400 | 12 585 | Proprietary |
| 142 | Gemini 3.1 Flash Lite Preview | 1 394 | 1 388–1 399 | 25 055 | Proprietary | |
| 143 | Grok 4.3 | xAI | 1 394 | 1 388–1 399 | 30 526 | Proprietary |
| 144 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 393 | 1 380–1 407 | 1 744 | Apache 2.0 |
| 145 | Qwen3.5-Flash | Alibaba Qwen | 1 393 | 1 387–1 398 | 23 656 | Proprietary |
| 146 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 392 | 1 382–1 403 | 2 811 | MIT |
| 147 | LongCat-Flash-Chat · chat | Meituan | 1 392 | 1 380–1 404 | 2 349 | MIT |
| 148 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 391 | 1 385–1 398 | 11 176 | Apache 2.0 |
| 149 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 391 | 1 371–1 411 | 783 | Proprietary |
| 150 | R1 0528 | DeepSeek | 1 390 | 1 380–1 401 | 3 115 | MIT |
| 151 | Grok 4.1 Fast | xAI | 1 390 | 1 385–1 395 | 17 584 | Proprietary |
| 152 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 389 | 1 382–1 396 | 7 153 | Apache 2.0 |
| 153 | GPT-5 в режиме high | OpenAI | 1 389 | 1 381–1 397 | 6 910 | Proprietary |
| 154 | Hunyuan T1 | Tencent | 1 385 | 1 365–1 404 | 898 | Proprietary |
| 155 | GPT-4.1 | OpenAI | 1 385 | 1 378–1 391 | 9 852 | Proprietary |
| 156 | Muse Glimmer 30B | Meta | 1 383 | 1 368–1 398 | 1 669 | Apache-2.0 |
| 157 | Claude Sonnet 4 · 20250514 | Anthropic | 1 380 | 1 373–1 387 | 7 658 | Proprietary |
| 158 | Solar Pro 4 | Upstage | 1 379 | 1 361–1 397 | 1 164 | Proprietary |
| 159 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 378 | 1 370–1 387 | 4 921 | Apache 2.0 |
| 160 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 378 | 1 369–1 386 | 4 925 | Apache 2.0 |
| 161 | o1 · 2024-12-17 | OpenAI | 1 378 | 1 368–1 388 | 3 571 | Proprietary |
| 162 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 373 | 1 365–1 381 | 6 126 | Proprietary |
| 163 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 373 | 1 365–1 380 | 6 349 | Proprietary | |
| 164 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 372 | 1 366–1 378 | 11 563 | Proprietary | |
| 165 | Grok 3 Mini в режиме high | xAI | 1 371 | 1 361–1 382 | 3 215 | Proprietary |
| 166 | o3 | OpenAI | 1 371 | 1 365–1 377 | 11 459 | Proprietary |
| 167 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 370 | 1 359–1 382 | 2 625 | Proprietary |
| 168 | GPT-5.4 Nano в режиме high | OpenAI | 1 366 | 1 361–1 372 | 25 378 | Proprietary |
| 169 | GLM 4.5 Air | Z.ai | 1 365 | 1 358–1 373 | 6 666 | MIT |
| 170 | Nemotron 3 Super | NVIDIA | 1 365 | 1 352–1 377 | 2 093 | NVIDIA Open Model |
| 171 | MiniMax M2.5 | MiniMax | 1 365 | 1 359–1 371 | 15 268 | Modified MIT |
| 172 | Hunyuan TurboS · 20250416 | Tencent | 1 364 | 1 349–1 379 | 1 514 | Proprietary |
| 173 | Mistral Medium 3 | Mistral AI | 1 360 | 1 352–1 368 | 5 899 | Proprietary |
| 174 | Qwen2.5 Max | Alibaba Qwen | 1 358 | 1 349–1 366 | 4 522 | Proprietary |
| 175 | Granite 4.2 30B | IBM Granite | 1 357 | 1 341–1 374 | 1 351 | Apache 2.0 |
| 176 | Trinity Large | Arcee AI | 1 356 | 1 350–1 363 | 10 870 | Apache 2.0 |
| 177 | GLM 4.6V | Z.ai | 1 356 | 1 333–1 379 | 672 | MIT |
| 178 | Grok 3 Mini | xAI | 1 356 | 1 346–1 365 | 4 021 | Proprietary |
| 179 | R1 | DeepSeek | 1 355 | 1 343–1 367 | 2 303 | MIT |
| 180 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 355 | 1 347–1 362 | 6 643 | Proprietary |
| 181 | GPT-5 Mini в режиме high | OpenAI | 1 354 | 1 346–1 362 | 5 823 | Proprietary |
| 182 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 353 | 1 342–1 364 | 2 852 | Apache 2.0 |
| 183 | DeepSeek V3 0324 | DeepSeek | 1 353 | 1 346–1 359 | 8 493 | MIT |
| 184 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 352 | 1 343–1 361 | 4 594 | Apache 2.0 |
| 185 | Kimi K2 0905 | Moonshot AI | 1 351 | 1 340–1 363 | 2 573 | Modified MIT |
| 186 | Step-1o Turbo | StepFun | 1 349 | 1 333–1 366 | 1 265 | Proprietary |
| 187 | GLM 4.7 Flash | Z.ai | 1 347 | 1 337–1 358 | 3 170 | MIT |
| 188 | Gemini 2.0 Flash | 1 344 | 1 337–1 352 | 6 500 | Proprietary | |
| 189 | o1 · preview | OpenAI | 1 344 | 1 334–1 354 | 4 578 | Proprietary |
| 190 | GPT-4.1 Mini | OpenAI | 1 343 | 1 336–1 351 | 7 053 | Proprietary |
| 191 | DeepSeek V3 | DeepSeek | 1 343 | 1 332–1 353 | 3 123 | DeepSeek |
| 192 | o3 Mini High | OpenAI | 1 343 | 1 331–1 355 | 2 160 | Proprietary |
| 193 | Hunyuan Large | Tencent | 1 341 | 1 314–1 369 | 371 | Proprietary |
| 194 | Nova 2 Lite | Amazon | 1 336 | 1 325–1 346 | 2 959 | Proprietary |
| 195 | MiniMax M2 | MiniMax | 1 335 | 1 320–1 349 | 1 612 | Apache 2.0 |
| 196 | Amazon Nova Experimental Chat · 10-09 | Amazon | 1 334 | 1 310–1 358 | 570 | Proprietary |
| 197 | Command A | Cohere | 1 334 | 1 327–1 340 | 10 605 | CC-BY-NC-4.0 |
| 198 | Gemma 3 27B | 1 333 | 1 325–1 340 | 6 620 | Gemma | |
| 199 | Mercury 2 | Inception Labs | 1 331 | 1 311–1 351 | 844 | Proprietary |
| 200 | INTELLECT-3 | Prime Intellect | 1 329 | 1 313–1 346 | 1 278 | MIT |
| 201 | o3 Mini | OpenAI | 1 328 | 1 321–1 334 | 9 246 | Proprietary |
| 202 | Mistral Small 3.2 24B | Mistral AI | 1 327 | 1 317–1 337 | 3 410 | Apache 2.0 |
| 203 | Qwen3 32B | Alibaba Qwen | 1 327 | 1 303–1 351 | 504 | Apache 2.0 |
| 204 | MiniMax M1 | MiniMax | 1 326 | 1 319–1 334 | 6 916 | Apache 2.0 |
| 205 | Step 3 | StepFun | 1 326 | 1 310–1 342 | 1 268 | Apache 2.0 |
| 206 | Kimi K2 0711 | Moonshot AI | 1 325 | 1 317–1 334 | 5 516 | Modified MIT |
| 207 | Qwen-Plus | Alibaba Qwen | 1 324 | 1 303–1 345 | 690 | Proprietary |
| 208 | Ling-flash-2.0 | inclusionAI | 1 323 | 1 307–1 339 | 1 258 | MIT |
| 209 | Gemini 2.0 Flash-Lite | 1 320 | 1 310–1 331 | 2 940 | Proprietary | |
| 210 | o1-mini | OpenAI | 1 320 | 1 313–1 328 | 7 913 | Proprietary |
| 211 | Hunyuan TurboS · 20250226 | Tencent | 1 320 | 1 288–1 352 | 287 | Proprietary |
| 212 | Trinity Large Thinking | Arcee AI | 1 320 | 1 313–1 327 | 12 330 | Apache 2.0 |
| 213 | gpt-oss-120b | OpenAI | 1 319 | 1 312–1 327 | 6 487 | Apache 2.0 |
| 214 | Granite 4.2 8B | IBM Granite | 1 319 | 1 301–1 337 | 1 281 | Apache 2.0 |
| 215 | Ring-flash-2.0 | inclusionAI | 1 319 | 1 303–1 335 | 1 332 | MIT |
| 216 | Nemotron 3.5 Lightning | NVIDIA | 1 317 | 1 307–1 328 | 3 731 | OpenMDW-1.1 |
| 217 | Gemma 3 12B | 1 317 | 1 289–1 345 | 371 | Gemma | |
| 218 | GLM-4-Plus · plus-0111 | Z.ai | 1 316 | 1 295–1 336 | 765 | Proprietary |
| 219 | o4 Mini | OpenAI | 1 314 | 1 308–1 321 | 8 571 | Proprietary |
| 220 | OLMo 3.1 32B Instruct | Ai2 | 1 313 | 1 302–1 325 | 2 791 | Apache 2.0 |
| 221 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 313 | 1 292–1 335 | 645 | Nvidia Open |
| 222 | Qwen3 30B A3B | Alibaba Qwen | 1 312 | 1 303–1 321 | 4 476 | Apache 2.0 |
| 223 | GPT-5 Nano в режиме high | OpenAI | 1 312 | 1 297–1 326 | 1 628 | Proprietary |
| 224 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 311 | 1 306–1 317 | 13 772 | Proprietary |
| 225 | Hunyuan Turbo | Tencent | 1 309 | 1 280–1 338 | 313 | Proprietary |
| 226 | QwQ 32B | Alibaba Qwen | 1 308 | 1 299–1 317 | 4 034 | Apache 2.0 |
| 227 | Gemini 1.5 Pro · 1.5-pro-002 | 1 308 | 1 301–1 315 | 8 372 | Proprietary | |
| 228 | Step-2 16k | StepFun | 1 306 | 1 285–1 327 | 693 | Proprietary |
| 229 | GLM 4.5V | Z.ai | 1 304 | 1 286–1 322 | 1 065 | MIT |
| 230 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 302 | 1 293–1 312 | 3 865 | NVIDIA Open Model |
| 231 | Hunyuan Standard · 2025-02-10 | Tencent | 1 301 | 1 274–1 327 | 432 | Proprietary |
| 232 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 301 | 1 284–1 318 | 1 037 | DeepSeek |
| 233 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 299 | 1 269–1 329 | 317 | Nvidia |
| 234 | Mistral Small 3.1 24B | Mistral AI | 1 299 | 1 291–1 306 | 6 324 | Apache 2.0 |
| 235 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 1 299 | 1 269–1 328 | 321 | Nvidia Open Model |
| 236 | Yi-Lightning | 01.AI | 1 297 | 1 286–1 307 | 3 634 | Proprietary |
| 237 | Granite 4.2 3B | IBM Granite | 1 295 | 1 276–1 314 | 1 248 | Apache 2.0 |
| 238 | OLMo 3 32B Think | Ai2 | 1 295 | 1 279–1 311 | 1 331 | Apache 2.0 |
| 239 | Magistral Medium | Mistral AI | 1 294 | 1 282–1 307 | 2 393 | Proprietary |
| 240 | Qwen2.5 Plus | Alibaba Qwen | 1 292 | 1 278–1 306 | 1 586 | Proprietary |
| 241 | Gemini 1.5 Pro · 1.5-pro-001 | 1 291 | 1 282–1 299 | 9 750 | Proprietary | |
| 242 | Athene V2 Chat | Nexusflow | 1 289 | 1 280–1 299 | 3 663 | NexusFlow |
| 243 | GPT-4o-mini (2024-07-18) | OpenAI | 1 289 | 1 282–1 296 | 8 986 | Proprietary |
| 244 | GPT-4o (2024-05-13) | OpenAI | 1 288 | 1 282–1 295 | 14 922 | Proprietary |
| 245 | Qwen Max | Alibaba Qwen | 1 288 | 1 277–1 300 | 2 440 | Qwen |
| 246 | GLM-4-Plus · plus | Z.ai | 1 286 | 1 276–1 296 | 3 992 | Proprietary |
| 247 | Gemini 1.5 Flash · 002 | 1 284 | 1 275–1 292 | 5 463 | Proprietary | |
| 248 | GPT-4o (2024-08-06) | OpenAI | 1 283 | 1 275–1 292 | 6 160 | Proprietary |
| 249 | GPT-4.1 Nano | OpenAI | 1 283 | 1 263–1 302 | 730 | Proprietary |
| 250 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 282 | 1 273–1 290 | 6 008 | Qwen |
| 251 | Llama 4 Maverick | Meta | 1 279 | 1 272–1 287 | 7 074 | Llama 4 |
| 252 | Hunyuan Large Vision | Tencent | 1 279 | 1 259–1 299 | 871 | Proprietary |
| 253 | Granite 4.1 8B | IBM Granite | 1 277 | 1 260–1 293 | 1 569 | Apache 2.0 |
| 254 | Grok 2 | xAI | 1 276 | 1 269–1 283 | 8 902 | Proprietary |
| 255 | Gemma 3n E4B | 1 276 | 1 265–1 287 | 2 876 | Gemma | |
| 256 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 275 | 1 267–1 283 | 10 574 | Proprietary |
| 257 | Gemma 3 4B | 1 273 | 1 246–1 300 | 444 | Gemma | |
| 258 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 273 | 1 263–1 283 | 3 499 | DeepSeek |
| 259 | OLMo 3.1 32B Think | Ai2 | 1 268 | 1 254–1 282 | 1 976 | Apache 2.0 |
| 260 | Grok 2 Mini | xAI | 1 266 | 1 259–1 274 | 7 048 | Proprietary |
| 261 | Mercury | Inception Labs | 1 266 | 1 240–1 292 | 509 | Proprietary |
| 262 | Llama 3.1 405B Instruct · bf16 | Meta | 1 266 | 1 258–1 274 | 5 486 | Llama 3.1 Community |
| 263 | Llama 4 Scout | Meta | 1 265 | 1 257–1 274 | 5 802 | Llama |
| 264 | Mistral Large | Mistral AI | 1 261 | 1 252–1 271 | 3 607 | MRL |
| 265 | Claude 3.5 Haiku | Anthropic | 1 260 | 1 254–1 266 | 10 948 | Proprietary |
| 266 | Llama 3.1 405B Instruct · fp8 | Meta | 1 259 | 1 252–1 267 | 8 047 | Llama 3.1 Community |
| 267 | Claude 3 Opus | Anthropic | 1 259 | 1 253–1 265 | 23 374 | Proprietary |
| 268 | Mistral Large 2407 | Mistral AI | 1 257 | 1 249–1 266 | 5 885 | Mistral Research |
| 269 | Llama 3.3 70B Instruct | Meta | 1 255 | 1 249–1 262 | 8 319 | Llama-3.3 |
| 270 | Nova Pro 1.0 | Amazon | 1 255 | 1 245–1 265 | 3 398 | Proprietary |
| 271 | GPT-4 Turbo | OpenAI | 1 254 | 1 246–1 262 | 11 559 | Proprietary |
| 272 | Gemini 1.5 Pro · advanced-0514 | 1 253 | 1 243–1 264 | 5 486 | Proprietary | |
| 273 | Gemini 1.5 Flash · 001 | 1 252 | 1 243–1 260 | 7 837 | Proprietary | |
| 274 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 251 | 1 232–1 269 | 804 | Apache 2.0 |
| 275 | gpt-oss-20b | OpenAI | 1 250 | 1 237–1 263 | 2 114 | Apache 2.0 |
| 276 | Mistral Small 3 | Mistral AI | 1 246 | 1 232–1 259 | 1 767 | Apache 2.0 |
| 277 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 244 | 1 235–1 252 | 9 653 | Proprietary |
| 278 | Llama 3.1 70B Instruct | Meta | 1 241 | 1 233–1 248 | 7 622 | Llama 3.1 Community |
| 279 | Granite 4.0 H Small | IBM Granite | 1 239 | 1 221–1 256 | 1 285 | Apache 2.0 |
| 280 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 238 | 1 221–1 256 | 1 049 | Llama 3.1 |
| 281 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 236 | 1 228–1 244 | 9 140 | Proprietary |
| 282 | Nova Lite 1.0 | Amazon | 1 234 | 1 223–1 244 | 2 784 | Proprietary |
| 283 | Athene 70B | Nexusflow | 1 232 | 1 220–1 244 | 2 123 | CC-BY-NC-4.0 |
| 284 | Gemma 2 27B | 1 230 | 1 224–1 237 | 9 918 | Gemma license | |
| 285 | Command R+ (08-2024) | Cohere | 1 230 | 1 215–1 245 | 1 369 | CC-BY-NC-4.0 |
| 286 | Aya Expanse 32B | Cohere | 1 227 | 1 218–1 237 | 4 165 | CC-BY-NC-4.0 |
| 287 | Gemma 2 9B IT SimPO | Princeton NLP | 1 226 | 1 207–1 245 | 841 | MIT |
| 288 | Nemotron-4 340B Instruct | NVIDIA | 1 224 | 1 211–1 238 | 2 130 | NVIDIA Open Model |
| 289 | Hunyuan Standard · 256k | Tencent | 1 224 | 1 198–1 251 | 414 | Proprietary |
| 290 | Llama 3.1 Tulu 3 70B | Ai2 | 1 224 | 1 199–1 249 | 463 | Llama 3.1 |
| 291 | Reka Core | Reka AI | 1 220 | 1 202–1 238 | 951 | Proprietary |
| 292 | DeepSeek Coder V2 | DeepSeek | 1 219 | 1 205–1 233 | 1 842 | DeepSeek License |
| 293 | Gemini 1.5 Flash-8B | 1 218 | 1 210–1 227 | 5 573 | Proprietary | |
| 294 | Phi 4 | Microsoft | 1 217 | 1 206–1 228 | 2 896 | MIT |
| 295 | Ministral 8B (2410) | Mistral AI | 1 212 | 1 192–1 232 | 720 | MRL |
| 296 | Claude 3 Sonnet | Anthropic | 1 211 | 1 202–1 220 | 11 870 | Proprietary |
| 297 | GLM-4 | Z.ai | 1 207 | 1 190–1 223 | 1 174 | Proprietary |
| 298 | Jamba 1.5 Large | AI21 Labs | 1 206 | 1 189–1 224 | 1 024 | Jamba Open |
| 299 | Nova Micro 1.0 | Amazon | 1 205 | 1 194–1 217 | 2 625 | Proprietary |
| 300 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 205 | 1 182–1 228 | 587 | 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.