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: медицина, здравоохранение», 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 | 1 518 | 1 489–1 546 | 435 | Proprietary |
| 2 | Claude Opus 5 в режиме max | Anthropic | 1 516 | 1 500–1 531 | 1 569 | Proprietary |
| 3 | Muse Spark 1.2 в режиме xhigh | Meta | 1 515 | 1 477–1 553 | 232 | Proprietary |
| 4 | Claude Opus 5 в режиме high | Anthropic | 1 509 | 1 498–1 520 | 3 233 | Proprietary |
| 5 | Claude Opus 4.6 в режиме high | Anthropic | 1 506 | 1 497–1 515 | 5 357 | Proprietary |
| 6 | Claude Opus 4.6 | Anthropic | 1 500 | 1 492–1 509 | 5 623 | Proprietary |
| 7 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 500 | 1 481–1 518 | 1 079 | Proprietary |
| 8 | Qwen3.5 Max | Alibaba Qwen | 1 499 | 1 484–1 515 | 1 566 | Proprietary |
| 9 | Gemini 3.8 Flash в режиме high | 1 497 | 1 469–1 526 | 416 | Proprietary | |
| 10 | Claude Opus 4.7 в режиме high | Anthropic | 1 497 | 1 488–1 507 | 4 498 | Proprietary |
| 11 | Gemini 3 Pro | 1 492 | 1 480–1 504 | 2 533 | Proprietary | |
| 12 | Claude Opus 4.7 | Anthropic | 1 491 | 1 481–1 500 | 4 528 | Proprietary |
| 13 | Gemini 3.5 Flash в режиме high | 1 490 | 1 478–1 502 | 2 831 | Proprietary | |
| 14 | Muse Spark 1.3 в режиме max | Meta | 1 489 | 1 457–1 521 | 331 | Proprietary |
| 15 | ERNIE 5.1 | Baidu | 1 488 | 1 476–1 500 | 2 685 | Proprietary |
| 16 | Gemini 3.1 Pro Preview | 1 485 | 1 477–1 493 | 7 936 | Proprietary | |
| 17 | GLM 5.3 в режиме max | Z.ai | 1 481 | 1 460–1 503 | 784 | MIT |
| 18 | Kimi K3 в режиме max | Moonshot AI | 1 481 | 1 465–1 497 | 1 415 | Kimi K3 license |
| 19 | Gemini 3.7 Flash в режиме high | 1 481 | 1 449–1 513 | 381 | Proprietary | |
| 20 | Muse Spark | Meta | 1 481 | 1 461–1 501 | 954 | Proprietary |
| 21 | Gemini 3.5 Flash в режиме medium | 1 481 | 1 468–1 493 | 2 637 | Proprietary | |
| 22 | Claude Fable 5 | Anthropic | 1 479 | 1 466–1 493 | 2 125 | Proprietary |
| 23 | Gemini 2.5 Pro | 1 476 | 1 469–1 483 | 7 778 | Proprietary | |
| 24 | MiMo-V2.5-Pro | Xiaomi | 1 474 | 1 464–1 483 | 4 342 | MIT |
| 25 | Gemini 3.6 Flash в режиме high | 1 473 | 1 459–1 487 | 1 907 | Proprietary | |
| 26 | ERNIE 5.0 · preview-1203 | Baidu | 1 473 | 1 450–1 496 | 676 | Proprietary |
| 27 | GLM 5.1 | Z.ai | 1 473 | 1 462–1 483 | 3 514 | MIT |
| 28 | Muse Spark 1.1 | Meta | 1 472 | 1 459–1 486 | 2 043 | Proprietary |
| 29 | Seed 2.0 Pro | ByteDance Seed | 1 471 | 1 462–1 480 | 5 444 | Proprietary |
| 30 | GLM 5.2 в режиме max | Z.ai | 1 471 | 1 459–1 483 | 2 652 | MIT |
| 31 | Qwen3.7 Max | Alibaba Qwen | 1 470 | 1 435–1 505 | 296 | Proprietary |
| 32 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 469 | 1 441–1 496 | 436 | MIT |
| 33 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 468 | 1 458–1 479 | 3 582 | MIT |
| 34 | Qwen3 Max · preview | Alibaba Qwen | 1 468 | 1 452–1 484 | 1 456 | Proprietary |
| 35 | Grok 3 | xAI | 1 467 | 1 452–1 482 | 1 607 | Proprietary |
| 36 | DeepSeek V3.1 Terminus · terminus-thinking | DeepSeek | 1 467 | 1 429–1 505 | 227 | MIT |
| 37 | Gemini 3 Flash Preview | 1 465 | 1 452–1 478 | 2 073 | Proprietary | |
| 38 | GPT-5.5 · 5.5 | OpenAI | 1 464 | 1 455–1 474 | 4 963 | Proprietary |
| 39 | GLM 4.7 | Z.ai | 1 464 | 1 443–1 484 | 803 | MIT |
| 40 | DeepSeek V4 Pro 0423 | DeepSeek | 1 464 | 1 453–1 474 | 3 766 | MIT |
| 41 | Qwen3.7 Plus | Alibaba Qwen | 1 464 | 1 452–1 476 | 2 732 | Proprietary |
| 42 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 462 | 1 424–1 500 | 230 | Proprietary |
| 43 | Qwen3.6 Max Preview | Alibaba Qwen | 1 462 | 1 430–1 493 | 356 | Proprietary |
| 44 | LongCat-Flash-Chat · chat | Meituan | 1 460 | 1 437–1 483 | 666 | MIT |
| 45 | ERNIE 5.0 · 0110 | Baidu | 1 460 | 1 448–1 472 | 2 379 | Proprietary |
| 46 | GPT-5.4 в режиме high | OpenAI | 1 459 | 1 450–1 469 | 4 457 | Proprietary |
| 47 | Claude Opus 4.8 в режиме high | Anthropic | 1 459 | 1 448–1 469 | 3 919 | Proprietary |
| 48 | Claude Opus 4.8 | Anthropic | 1 458 | 1 448–1 469 | 3 891 | Proprietary |
| 49 | GPT-5.1 в режиме high | OpenAI | 1 458 | 1 446–1 470 | 2 485 | Proprietary |
| 50 | GLM 4.6 | Z.ai | 1 457 | 1 443–1 470 | 1 874 | MIT |
| 51 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 456 | 1 439–1 474 | 1 176 | Apache 2.0 |
| 52 | GPT-5.5 в режиме high | OpenAI | 1 456 | 1 447–1 465 | 4 865 | Proprietary |
| 53 | Nemotron 3 Ultra | NVIDIA | 1 456 | 1 436–1 477 | 800 | OpenMDW-1.1 |
| 54 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 456 | 1 430–1 482 | 509 | Apache 2.0 |
| 55 | Hy3 | Tencent | 1 456 | 1 429–1 482 | 520 | Apache 2.0 |
| 56 | GLM 4.5 | Z.ai | 1 455 | 1 439–1 471 | 1 379 | MIT |
| 57 | Kimi K2.6 | Moonshot AI | 1 455 | 1 443–1 467 | 2 702 | Modified MIT |
| 58 | GLM 5 | Z.ai | 1 454 | 1 441–1 468 | 2 014 | MIT |
| 59 | Grok 4.20 · beta-0309-reasoning | xAI | 1 454 | 1 445–1 464 | 4 557 | Proprietary |
| 60 | Claude Sonnet 4.6 | Anthropic | 1 454 | 1 445–1 463 | 5 023 | Proprietary |
| 61 | Qwen3.5 397B A17B | Alibaba Qwen | 1 453 | 1 445–1 462 | 5 749 | Apache 2.0 |
| 62 | Grok 4.1 · 4.1-thinking | xAI | 1 453 | 1 444–1 463 | 4 274 | Proprietary |
| 63 | ChatGPT-4o (latest) | OpenAI | 1 453 | 1 444–1 462 | 4 503 | Proprietary |
| 64 | Hunyuan T1 | Tencent | 1 452 | 1 415–1 490 | 260 | Proprietary |
| 65 | DeepSeek V3.1 Terminus · terminus | DeepSeek | 1 452 | 1 414–1 489 | 259 | MIT |
| 66 | Grok 4.20 Multi-Agent | xAI | 1 451 | 1 442–1 461 | 4 478 | Proprietary |
| 67 | Grok 4.20 · beta1 | xAI | 1 451 | 1 437–1 465 | 1 927 | Proprietary |
| 68 | Mistral Medium 3.1 | Mistral AI | 1 450 | 1 442–1 458 | 5 815 | Proprietary |
| 69 | GLM 5.3 Flash | Z.ai | 1 449 | 1 427–1 472 | 650 | MIT |
| 70 | DeepSeek V4 Flash 0423 | DeepSeek | 1 449 | 1 438–1 460 | 3 430 | MIT |
| 71 | Kimi K2.5 · thinking | Moonshot AI | 1 449 | 1 440–1 458 | 4 821 | Modified MIT |
| 72 | o3 | OpenAI | 1 448 | 1 437–1 459 | 3 320 | Proprietary |
| 73 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 448 | 1 433–1 462 | 1 683 | Proprietary |
| 74 | Grok 4.5 | xAI | 1 447 | 1 434–1 460 | 2 202 | Proprietary |
| 75 | Mistral Large 3 2512 | Mistral AI | 1 447 | 1 438–1 456 | 4 781 | Apache 2.0 |
| 76 | Gemma 4 31B | 1 446 | 1 418–1 474 | 410 | Apache 2.0 | |
| 77 | MiniMax M3 | MiniMax | 1 446 | 1 435–1 457 | 3 284 | MiniMax Community License |
| 78 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 446 | 1 422–1 470 | 572 | MIT |
| 79 | GPT-6 Astra в режиме max | OpenAI | 1 446 | 1 406–1 485 | 199 | Proprietary |
| 80 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 445 | 1 437–1 454 | 5 713 | Apache 2.0 |
| 81 | Qwen3.6 Plus | Alibaba Qwen | 1 445 | 1 434–1 456 | 3 227 | Proprietary |
| 82 | Gemini 3 Flash Preview в режиме minimal | 1 445 | 1 437–1 453 | 6 254 | Proprietary | |
| 83 | R1 0528 | DeepSeek | 1 444 | 1 427–1 462 | 1 227 | MIT |
| 84 | Claude Sonnet 5 в режиме high | Anthropic | 1 444 | 1 432–1 456 | 2 549 | Proprietary |
| 85 | Grok 4.1 · 4.1 | xAI | 1 444 | 1 435–1 454 | 4 357 | Proprietary |
| 86 | GPT-5.1 | OpenAI | 1 444 | 1 432–1 456 | 2 628 | Proprietary |
| 87 | GPT-5.2 Chat | OpenAI | 1 444 | 1 431–1 456 | 2 456 | Proprietary |
| 88 | Claude Opus 4.5 | Anthropic | 1 442 | 1 433–1 451 | 4 683 | Proprietary |
| 89 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 442 | 1 419–1 464 | 695 | MIT |
| 90 | Claude Sonnet 4.5 | Anthropic | 1 441 | 1 433–1 450 | 5 131 | Proprietary |
| 91 | GPT-5.4 | OpenAI | 1 441 | 1 432–1 451 | 4 676 | Proprietary |
| 92 | Gemini 3.5 Flash Lite | 1 441 | 1 428–1 455 | 1 955 | Proprietary | |
| 93 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 440 | 1 429–1 451 | 3 490 | MIT |
| 94 | ERNIE 5.0 · preview-1022 | Baidu | 1 439 | 1 402–1 475 | 241 | Proprietary |
| 95 | Inkling | Thinking Machines Lab | 1 439 | 1 424–1 453 | 1 762 | Apache 2.0 |
| 96 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 437 | 1 426–1 448 | 2 994 | MIT |
| 97 | MiMo-V2 Pro | Xiaomi | 1 437 | 1 422–1 452 | 1 708 | Proprietary |
| 98 | Grok 4 | xAI | 1 434 | 1 421–1 447 | 2 229 | Proprietary |
| 99 | MiMo-V2 Omni | Xiaomi | 1 434 | 1 418–1 451 | 1 331 | Proprietary |
| 100 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 433 | 1 419–1 447 | 1 981 | Proprietary |
| 101 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 432 | 1 408–1 457 | 622 | Apache 2.0 |
| 102 | Qwen3.8 27B | Alibaba Qwen | 1 432 | 1 409–1 455 | 699 | Apache 2.0 |
| 103 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 431 | 1 417–1 444 | 2 067 | Proprietary |
| 104 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 431 | 1 412–1 450 | 940 | MIT |
| 105 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 430 | 1 422–1 439 | 5 171 | Proprietary |
| 106 | MiMo-V2.5 | Xiaomi | 1 430 | 1 418–1 441 | 3 159 | MIT |
| 107 | Grok 4.6 в режиме high | xAI | 1 429 | 1 411–1 448 | 1 079 | Proprietary |
| 108 | Gemini 3.1 Flash Lite Preview | 1 429 | 1 419–1 438 | 4 498 | Proprietary | |
| 109 | Gemini 2.5 Flash · flash | 1 428 | 1 421–1 435 | 7 667 | Proprietary | |
| 110 | GLM 5V Turbo | Z.ai | 1 428 | 1 405–1 451 | 646 | Proprietary |
| 111 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 428 | 1 404–1 452 | 580 | MIT |
| 112 | Claude Opus 4.5 в режиме high | Anthropic | 1 427 | 1 415–1 440 | 2 319 | Proprietary |
| 113 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 427 | 1 414–1 441 | 2 075 | Apache 2.0 |
| 114 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 427 | 1 389–1 464 | 215 | Proprietary |
| 115 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 426 | 1 412–1 440 | 1 971 | Proprietary |
| 116 | Gemma 4 26B A4B | 1 425 | 1 395–1 455 | 369 | Apache 2.0 | |
| 117 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 425 | 1 412–1 438 | 2 174 | Apache 2.0 |
| 118 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 425 | 1 410–1 439 | 1 635 | Proprietary | |
| 119 | GPT-5.2 | OpenAI | 1 424 | 1 415–1 432 | 5 760 | Proprietary |
| 120 | Qwen3.5-27B | Alibaba Qwen | 1 423 | 1 409–1 436 | 1 928 | Apache 2.0 |
| 121 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 421 | 1 410–1 433 | 2 709 | MIT |
| 122 | GPT-5.2 в режиме high | OpenAI | 1 420 | 1 409–1 431 | 3 341 | Proprietary |
| 123 | Grok 4 Fast · reasoning | xAI | 1 420 | 1 400–1 440 | 854 | Proprietary |
| 124 | Hunyuan TurboS | Tencent | 1 419 | 1 394–1 444 | 583 | Proprietary |
| 125 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 418 | 1 390–1 446 | 431 | Apache 2.0 |
| 126 | Kimi K2 Thinking | Moonshot AI | 1 418 | 1 408–1 427 | 3 849 | Modified MIT |
| 127 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 416 | 1 405–1 426 | 3 156 | MIT |
| 128 | GPT-5 | OpenAI | 1 414 | 1 399–1 428 | 1 677 | Proprietary |
| 129 | GPT-5 в режиме high | OpenAI | 1 414 | 1 399–1 428 | 1 777 | Proprietary |
| 130 | Inkling Small | Thinking Machines Lab | 1 414 | 1 397–1 430 | 1 414 | Apache 2.0 |
| 131 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 413 | 1 402–1 425 | 2 640 | Proprietary |
| 132 | Grok 4.1 Fast | xAI | 1 413 | 1 403–1 423 | 3 659 | Proprietary |
| 133 | GPT-5.5 · 5.5-instant | OpenAI | 1 413 | 1 399–1 427 | 1 967 | Proprietary |
| 134 | Claude Opus 4.1 · 20250805 | Anthropic | 1 413 | 1 403–1 422 | 4 306 | Proprietary |
| 135 | Grok 4 Fast · chat | xAI | 1 412 | 1 382–1 443 | 378 | Proprietary |
| 136 | Mistral Medium 3.5 | Mistral AI | 1 412 | 1 391–1 433 | 803 | Modified MIT |
| 137 | Hy3 preview | Tencent | 1 412 | 1 383–1 440 | 459 | tencent-hunyuan-community |
| 138 | Kimi K2.5 · instant | Moonshot AI | 1 411 | 1 386–1 436 | 533 | Modified MIT |
| 139 | MiniMax M2.7 | MiniMax | 1 411 | 1 401–1 420 | 4 716 | Modified MIT |
| 140 | GPT-4.5 Preview | OpenAI | 1 410 | 1 386–1 435 | 589 | Proprietary |
| 141 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 409 | 1 385–1 432 | 636 | Proprietary |
| 142 | Step 3.5 Flash | StepFun | 1 409 | 1 399–1 418 | 3 998 | Apache 2.0 |
| 143 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 407 | 1 395–1 420 | 2 315 | Apache 2.0 |
| 144 | Qwen3.5-Flash | Alibaba Qwen | 1 407 | 1 397–1 417 | 4 078 | Proprietary |
| 145 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 407 | 1 379–1 435 | 445 | Proprietary |
| 146 | Muse Glimmer 30B | Meta | 1 405 | 1 367–1 443 | 247 | Apache-2.0 |
| 147 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 405 | 1 383–1 427 | 687 | MIT |
| 148 | GLM 4.5 Air | Z.ai | 1 401 | 1 387–1 416 | 1 633 | MIT |
| 149 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 397 | 1 360–1 434 | 236 | Proprietary |
| 150 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 397 | 1 383–1 410 | 2 131 | Proprietary |
| 151 | GPT-5.4 Mini в режиме high | OpenAI | 1 395 | 1 385–1 405 | 4 427 | Proprietary |
| 152 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 393 | 1 371–1 415 | 710 | Apache 2.0 |
| 153 | Claude Haiku 4.5 | Anthropic | 1 393 | 1 385–1 400 | 8 562 | Proprietary |
| 154 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 392 | 1 375–1 408 | 1 323 | Apache 2.0 |
| 155 | Gemma 3 27B | 1 391 | 1 378–1 404 | 2 337 | Gemma | |
| 156 | Grok 3 Mini в режиме high | xAI | 1 390 | 1 371–1 410 | 950 | Proprietary |
| 157 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 390 | 1 378–1 402 | 2 485 | Proprietary | |
| 158 | GPT-5.3 Chat | OpenAI | 1 390 | 1 377–1 402 | 2 451 | Proprietary |
| 159 | GLM-4-Plus · plus-0111 | Z.ai | 1 389 | 1 356–1 423 | 293 | Proprietary |
| 160 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 388 | 1 375–1 402 | 1 962 | Proprietary | |
| 161 | gpt-oss-120b | OpenAI | 1 387 | 1 372–1 403 | 1 615 | Apache 2.0 |
| 162 | Nova 2 Lite | Amazon | 1 387 | 1 365–1 409 | 708 | Proprietary |
| 163 | MiniMax M2.1 | MiniMax | 1 386 | 1 368–1 404 | 1 117 | MIT |
| 164 | Nemotron 3 Super | NVIDIA | 1 385 | 1 359–1 411 | 481 | NVIDIA Open Model |
| 165 | Grok 4.3 | xAI | 1 384 | 1 375–1 394 | 4 974 | Proprietary |
| 166 | Kimi K2 0905 | Moonshot AI | 1 383 | 1 360–1 405 | 712 | Modified MIT |
| 167 | DeepSeek V3 0324 | DeepSeek | 1 382 | 1 371–1 394 | 2 647 | MIT |
| 168 | Qwen2.5 Max | Alibaba Qwen | 1 382 | 1 367–1 397 | 1 589 | Proprietary |
| 169 | Mistral Medium 3 | Mistral AI | 1 380 | 1 366–1 393 | 1 971 | Proprietary |
| 170 | GPT-4.1 | OpenAI | 1 379 | 1 368–1 391 | 2 767 | Proprietary |
| 171 | GPT-5 Mini в режиме high | OpenAI | 1 378 | 1 362–1 394 | 1 418 | Proprietary |
| 172 | GPT-5.4 Nano в режиме high | OpenAI | 1 371 | 1 361–1 381 | 4 436 | Proprietary |
| 173 | R1 | DeepSeek | 1 370 | 1 349–1 391 | 830 | MIT |
| 174 | Mercury 2 | Inception Labs | 1 369 | 1 331–1 406 | 225 | Proprietary |
| 175 | Gemini 2.0 Flash | 1 368 | 1 356–1 381 | 2 296 | Proprietary | |
| 176 | Qwen3 32B | Alibaba Qwen | 1 368 | 1 330–1 406 | 212 | Apache 2.0 |
| 177 | MiniMax M2 | MiniMax | 1 368 | 1 335–1 401 | 316 | Apache 2.0 |
| 178 | Kimi K2 0711 | Moonshot AI | 1 367 | 1 352–1 382 | 1 687 | Modified MIT |
| 179 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 367 | 1 326–1 408 | 195 | Nvidia Open |
| 180 | Grok 3 Mini | xAI | 1 365 | 1 349–1 382 | 1 364 | Proprietary |
| 181 | Claude Opus 4 · 20250514 | Anthropic | 1 365 | 1 353–1 377 | 2 600 | Proprietary |
| 182 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 363 | 1 348–1 378 | 1 530 | Apache 2.0 |
| 183 | Ling-flash-2.0 | inclusionAI | 1 361 | 1 331–1 392 | 349 | MIT |
| 184 | o4 Mini | OpenAI | 1 361 | 1 349–1 374 | 2 480 | Proprietary |
| 185 | Step 3 | StepFun | 1 360 | 1 329–1 390 | 349 | Apache 2.0 |
| 186 | MiniMax M2.5 | MiniMax | 1 359 | 1 347–1 370 | 3 064 | Modified MIT |
| 187 | MiniMax M1 | MiniMax | 1 355 | 1 342–1 368 | 2 081 | Apache 2.0 |
| 188 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 355 | 1 342–1 368 | 2 219 | Proprietary |
| 189 | Trinity Large Thinking | Arcee AI | 1 354 | 1 341–1 368 | 2 180 | Apache 2.0 |
| 190 | o1 · 2024-12-17 | OpenAI | 1 354 | 1 337–1 371 | 1 273 | Proprietary |
| 191 | Step-2 16k | StepFun | 1 354 | 1 319–1 389 | 265 | Proprietary |
| 192 | Qwen-Plus | Alibaba Qwen | 1 349 | 1 318–1 381 | 315 | Proprietary |
| 193 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 349 | 1 330–1 369 | 980 | NVIDIA Open Model |
| 194 | Step-1o Turbo | StepFun | 1 349 | 1 324–1 373 | 597 | Proprietary |
| 195 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 347 | 1 331–1 363 | 1 482 | Apache 2.0 |
| 196 | GLM 4.5V | Z.ai | 1 347 | 1 310–1 384 | 234 | MIT |
| 197 | Trinity Large | Arcee AI | 1 346 | 1 333–1 360 | 2 214 | Apache 2.0 |
| 198 | Mistral Small 3.2 24B | Mistral AI | 1 345 | 1 328–1 363 | 1 129 | Apache 2.0 |
| 199 | Gemini 2.0 Flash-Lite | 1 345 | 1 326–1 364 | 1 055 | Proprietary | |
| 200 | DeepSeek V3 | DeepSeek | 1 345 | 1 326–1 363 | 1 060 | DeepSeek |
| 201 | GLM 4.7 Flash | Z.ai | 1 343 | 1 322–1 364 | 799 | MIT |
| 202 | Gemma 3 4B | 1 341 | 1 303–1 380 | 197 | Gemma | |
| 203 | Hunyuan Large | Tencent | 1 339 | 1 301–1 378 | 212 | Proprietary |
| 204 | QwQ 32B | Alibaba Qwen | 1 339 | 1 323–1 355 | 1 348 | Apache 2.0 |
| 205 | Granite 4.2 30B | IBM Granite | 1 338 | 1 297–1 380 | 223 | Apache 2.0 |
| 206 | GPT-5 Nano в режиме high | OpenAI | 1 335 | 1 308–1 362 | 474 | Proprietary |
| 207 | Gemma 3 12B | 1 334 | 1 287–1 381 | 155 | Gemma | |
| 208 | Command A | Cohere | 1 334 | 1 323–1 345 | 3 043 | CC-BY-NC-4.0 |
| 209 | Ring-flash-2.0 | inclusionAI | 1 331 | 1 300–1 361 | 377 | MIT |
| 210 | o3 Mini High | OpenAI | 1 330 | 1 309–1 351 | 824 | Proprietary |
| 211 | Claude Sonnet 4 · 20250514 | Anthropic | 1 329 | 1 316–1 342 | 2 363 | Proprietary |
| 212 | Granite 4.2 8B | IBM Granite | 1 327 | 1 286–1 368 | 227 | Apache 2.0 |
| 213 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 326 | 1 313–1 340 | 2 120 | Proprietary |
| 214 | GPT-4.1 Mini | OpenAI | 1 324 | 1 311–1 337 | 2 314 | Proprietary |
| 215 | Gemma 3n E4B | 1 321 | 1 305–1 338 | 1 322 | Gemma | |
| 216 | INTELLECT-3 | Prime Intellect | 1 321 | 1 287–1 354 | 308 | MIT |
| 217 | Qwen3 30B A3B | Alibaba Qwen | 1 317 | 1 302–1 332 | 1 614 | Apache 2.0 |
| 218 | OLMo 3 32B Think | Ai2 | 1 316 | 1 280–1 352 | 306 | Apache 2.0 |
| 219 | Hunyuan Standard | Tencent | 1 309 | 1 272–1 345 | 247 | Proprietary |
| 220 | gpt-oss-20b | OpenAI | 1 309 | 1 283–1 334 | 563 | Apache 2.0 |
| 221 | o1 · preview | OpenAI | 1 306 | 1 291–1 322 | 1 699 | Proprietary |
| 222 | Grok 2 | xAI | 1 306 | 1 294–1 318 | 3 206 | Proprietary |
| 223 | Yi-Lightning | 01.AI | 1 305 | 1 289–1 322 | 1 451 | Proprietary |
| 224 | Gemini 1.5 Pro · 1.5-pro-002 | 1 305 | 1 293–1 318 | 2 686 | Proprietary | |
| 225 | Nemotron 3.5 Lightning | NVIDIA | 1 304 | 1 277–1 331 | 597 | OpenMDW-1.1 |
| 226 | Qwen2.5 Plus | Alibaba Qwen | 1 301 | 1 275–1 327 | 474 | Proprietary |
| 227 | Granite 4.1 8B | IBM Granite | 1 301 | 1 264–1 337 | 306 | Apache 2.0 |
| 228 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 300 | 1 270–1 331 | 392 | Llama 3.1 |
| 229 | o3 Mini | OpenAI | 1 299 | 1 287–1 310 | 3 037 | Proprietary |
| 230 | Grok 2 Mini | xAI | 1 297 | 1 284–1 310 | 2 609 | Proprietary |
| 231 | o1-mini | OpenAI | 1 296 | 1 283–1 309 | 2 639 | Proprietary |
| 232 | Athene V2 Chat | Nexusflow | 1 295 | 1 278–1 312 | 1 198 | NexusFlow |
| 233 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 291 | 1 259–1 324 | 291 | DeepSeek |
| 234 | Gemini 1.5 Flash · 002 | 1 289 | 1 274–1 304 | 1 696 | Proprietary | |
| 235 | GLM-4-Plus · plus | Z.ai | 1 289 | 1 273–1 305 | 1 513 | Proprietary |
| 236 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 288 | 1 275–1 302 | 2 246 | Proprietary |
| 237 | Llama 3.1 405B Instruct · bf16 | Meta | 1 286 | 1 272–1 300 | 1 899 | Llama 3.1 Community |
| 238 | OLMo 3.1 32B Instruct | Ai2 | 1 285 | 1 262–1 308 | 756 | Apache 2.0 |
| 239 | GPT-4o-mini (2024-07-18) | OpenAI | 1 281 | 1 270–1 292 | 3 439 | Proprietary |
| 240 | Granite 4.2 3B | IBM Granite | 1 278 | 1 231–1 325 | 208 | Apache 2.0 |
| 241 | Reka Core | Reka AI | 1 278 | 1 248–1 307 | 364 | Proprietary |
| 242 | GPT-4o (2024-05-13) | OpenAI | 1 277 | 1 267–1 288 | 5 984 | Proprietary |
| 243 | GPT-4.1 Nano | OpenAI | 1 276 | 1 244–1 308 | 302 | Proprietary |
| 244 | Llama 4 Scout | Meta | 1 276 | 1 261–1 290 | 1 821 | Llama |
| 245 | Llama 4 Maverick | Meta | 1 275 | 1 262–1 288 | 2 301 | Llama 4 |
| 246 | Llama 3.3 70B Instruct | Meta | 1 274 | 1 262–1 286 | 2 733 | Llama-3.3 |
| 247 | Athene 70B | Nexusflow | 1 272 | 1 253–1 291 | 1 039 | CC-BY-NC-4.0 |
| 248 | Llama 3.1 405B Instruct · fp8 | Meta | 1 270 | 1 258–1 282 | 3 064 | Llama 3.1 Community |
| 249 | Mistral Small 3.1 24B | Mistral AI | 1 267 | 1 252–1 281 | 1 943 | Apache 2.0 |
| 250 | OLMo 3.1 32B Think | Ai2 | 1 261 | 1 235–1 288 | 574 | Apache 2.0 |
| 251 | Nova Pro 1.0 | Amazon | 1 260 | 1 242–1 277 | 1 096 | Proprietary |
| 252 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 259 | 1 241–1 276 | 1 214 | DeepSeek |
| 253 | Qwen Max | Alibaba Qwen | 1 258 | 1 240–1 277 | 933 | Qwen |
| 254 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 258 | 1 244–1 272 | 2 250 | Proprietary |
| 255 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 258 | 1 244–1 272 | 1 956 | Qwen |
| 256 | Llama 3.1 70B Instruct | Meta | 1 256 | 1 244–1 269 | 2 821 | Llama 3.1 Community |
| 257 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 256 | 1 246–1 266 | 4 534 | Proprietary |
| 258 | Mistral Large | Mistral AI | 1 254 | 1 237–1 271 | 1 230 | MRL |
| 259 | GPT-4o (2024-08-06) | OpenAI | 1 251 | 1 237–1 265 | 2 174 | Proprietary |
| 260 | Gemini 1.5 Pro · advanced-0514 | 1 249 | 1 234–1 263 | 2 701 | Proprietary | |
| 261 | Magistral Medium | Mistral AI | 1 245 | 1 222–1 269 | 777 | Proprietary |
| 262 | Mistral Large 2407 | Mistral AI | 1 245 | 1 232–1 258 | 2 423 | Mistral Research |
| 263 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 244 | 1 233–1 255 | 4 460 | Proprietary |
| 264 | GPT-4 Turbo | OpenAI | 1 239 | 1 227–1 250 | 5 046 | Proprietary |
| 265 | Command R+ (08-2024) | Cohere | 1 237 | 1 213–1 262 | 547 | CC-BY-NC-4.0 |
| 266 | Hunyuan Large Vision | Tencent | 1 231 | 1 198–1 264 | 365 | Proprietary |
| 267 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 231 | 1 219–1 243 | 5 155 | Proprietary |
| 268 | Claude 3 Opus | Anthropic | 1 231 | 1 221–1 240 | 10 327 | Proprietary |
| 269 | Claude 3.5 Haiku | Anthropic | 1 227 | 1 216–1 238 | 3 635 | Proprietary |
| 270 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 224 | 1 212–1 235 | 5 514 | Proprietary |
| 271 | Aya Expanse 32B | Cohere | 1 222 | 1 205–1 238 | 1 358 | CC-BY-NC-4.0 |
| 272 | Gemma 2 9B IT SimPO | Princeton NLP | 1 221 | 1 195–1 247 | 516 | MIT |
| 273 | Granite 4.0 H Small | IBM Granite | 1 220 | 1 178–1 261 | 253 | Apache 2.0 |
| 274 | Mistral Small 3 | Mistral AI | 1 215 | 1 191–1 239 | 654 | Apache 2.0 |
| 275 | Llama 3 70B Instruct | Meta | 1 213 | 1 202–1 225 | 8 252 | Llama 3 Community |
| 276 | Gemini 1.5 Flash-8B | 1 213 | 1 198–1 228 | 1 813 | Proprietary | |
| 277 | Nova Lite 1.0 | Amazon | 1 212 | 1 192–1 232 | 872 | Proprietary |
| 278 | Gemini 1.5 Pro · 1.5-pro-001 | 1 211 | 1 199–1 224 | 4 418 | Proprietary | |
| 279 | Command R (08-2024) | Cohere | 1 208 | 1 182–1 233 | 500 | CC-BY-NC-4.0 |
| 280 | Nova Micro 1.0 | Amazon | 1 207 | 1 188–1 226 | 952 | Proprietary |
| 281 | Command R+ | Cohere | 1 203 | 1 190–1 215 | 4 198 | CC-BY-NC-4.0 |
| 282 | GLM-4 | Z.ai | 1 202 | 1 178–1 227 | 554 | Proprietary |
| 283 | Aya Expanse 8B | Cohere | 1 202 | 1 175–1 229 | 470 | CC-BY-NC-4.0 |
| 284 | Phi 4 | Microsoft | 1 201 | 1 182–1 221 | 1 037 | MIT |
| 285 | Reka Flash (2024-09) | Reka AI | 1 199 | 1 172–1 227 | 390 | Proprietary |
| 286 | Gemma 2 27B | 1 197 | 1 186–1 208 | 3 862 | Gemma license | |
| 287 | Jamba 1.5 Large | AI21 Labs | 1 196 | 1 168–1 223 | 474 | Jamba Open |
| 288 | Nemotron-4 340B Instruct | NVIDIA | 1 194 | 1 175–1 213 | 1 069 | NVIDIA Open Model |
| 289 | Claude 3 Sonnet | Anthropic | 1 188 | 1 175–1 200 | 5 941 | Proprietary |
| 290 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 186 | 1 146–1 225 | 213 | Llama 3.1 |
| 291 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 186 | 1 149–1 222 | 221 | Apache 2.0 |
| 292 | Gemini 1.5 Flash · 001 | 1 186 | 1 173–1 199 | 3 461 | Proprietary | |
| 293 | OLMo 2 32B Instruct | Ai2 | 1 181 | 1 137–1 225 | 204 | Apache-2.0 |
| 294 | Llama 3.1 8B Instruct | Meta | 1 181 | 1 168–1 194 | 2 564 | Llama 3.1 Community |
| 295 | Jamba 1.5 Mini | AI21 Labs | 1 180 | 1 154–1 207 | 486 | Jamba Open |
| 296 | Ministral 8B (2410) | Mistral AI | 1 179 | 1 146–1 213 | 280 | MRL |
| 297 | Qwen2 72B Instruct | Alibaba Qwen | 1 175 | 1 160–1 190 | 2 067 | Qianwen LICENSE |
| 298 | Gemma 2 9B | 1 165 | 1 152–1 177 | 2 812 | Gemma license | |
| 299 | Claude 3 Haiku | Anthropic | 1 164 | 1 153–1 176 | 6 385 | Proprietary |
| 300 | Yi-1.5 34B Chat | 01.AI | 1 159 | 1 141–1 176 | 1 330 | Apache-2.0 |
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.