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 «многошаговые диалоги», 300 строк. Оценка и интервал — из источника; интервал не равен диапазону мест.
| Место | Модель | Разработчик | Оценка | Интервал | Голоса | Лицензия |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 4.6 в режиме high | Anthropic | 1 512 | 1 505–1 518 | 12 295 | Proprietary |
| 2 | Gemini 3.8 Flash в режиме high | 1 510 | 1 488–1 531 | 784 | Proprietary | |
| 3 | Muse Spark 1.2 в режиме xhigh | Meta | 1 508 | 1 482–1 535 | 512 | Proprietary |
| 4 | Claude Opus 4.6 | Anthropic | 1 506 | 1 500–1 512 | 13 488 | Proprietary |
| 5 | Claude Opus 4.7 в режиме high | Anthropic | 1 504 | 1 497–1 511 | 10 589 | Proprietary |
| 6 | Claude Fable 5 | Anthropic | 1 504 | 1 495–1 513 | 5 048 | Proprietary |
| 7 | Claude Opus 4.7 | Anthropic | 1 504 | 1 497–1 511 | 10 890 | Proprietary |
| 8 | Gemini 3.7 Flash в режиме high | 1 498 | 1 477–1 518 | 883 | Proprietary | |
| 9 | Claude Opus 5 в режиме max | Anthropic | 1 497 | 1 486–1 509 | 3 196 | Proprietary |
| 10 | Claude Fable 5.1 в режиме max | Anthropic | 1 495 | 1 474–1 516 | 807 | Proprietary |
| 11 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 492 | 1 480–1 504 | 2 618 | Proprietary |
| 12 | Claude Opus 5 в режиме high | Anthropic | 1 491 | 1 483–1 500 | 6 842 | Proprietary |
| 13 | Muse Spark 1.3 в режиме max | Meta | 1 489 | 1 467–1 511 | 760 | Proprietary |
| 14 | Gemini 3.1 Pro Preview | 1 485 | 1 480–1 491 | 18 519 | Proprietary | |
| 15 | Muse Spark 1.1 | Meta | 1 485 | 1 475–1 495 | 4 320 | Proprietary |
| 16 | Gemini 3.5 Flash в режиме high | 1 484 | 1 476–1 492 | 6 491 | Proprietary | |
| 17 | Gemini 3 Pro | 1 483 | 1 476–1 491 | 6 618 | Proprietary | |
| 18 | GLM 5.3 в режиме max | Z.ai | 1 483 | 1 468–1 498 | 1 684 | MIT |
| 19 | Kimi K3 в режиме max | Moonshot AI | 1 483 | 1 472–1 493 | 3 311 | Kimi K3 license |
| 20 | GPT-5.4 в режиме high | OpenAI | 1 482 | 1 475–1 489 | 11 259 | Proprietary |
| 21 | Qwen3.7 Max | Alibaba Qwen | 1 480 | 1 457–1 503 | 658 | Proprietary |
| 22 | Qwen3.5 Max | Alibaba Qwen | 1 479 | 1 469–1 489 | 3 644 | Proprietary |
| 23 | Gemini 3.6 Flash в режиме high | 1 479 | 1 469–1 488 | 4 357 | Proprietary | |
| 24 | GPT-5.5 в режиме high | OpenAI | 1 478 | 1 471–1 484 | 11 169 | Proprietary |
| 25 | Muse Spark | Meta | 1 475 | 1 462–1 489 | 2 179 | Proprietary |
| 26 | Claude Opus 4.8 в режиме high | Anthropic | 1 475 | 1 467–1 482 | 9 353 | Proprietary |
| 27 | Claude Opus 4.8 | Anthropic | 1 474 | 1 467–1 481 | 9 603 | Proprietary |
| 28 | Gemini 3.5 Flash в режиме medium | 1 474 | 1 465–1 482 | 6 091 | Proprietary | |
| 29 | MiMo-V2.5-Pro | Xiaomi | 1 474 | 1 467–1 481 | 10 597 | MIT |
| 30 | Gemini 3 Flash Preview | 1 472 | 1 463–1 481 | 5 217 | Proprietary | |
| 31 | ERNIE 5.1 | Baidu | 1 471 | 1 463–1 480 | 6 412 | Proprietary |
| 32 | GLM 5.1 | Z.ai | 1 471 | 1 464–1 479 | 7 905 | MIT |
| 33 | GLM 5.3 Flash | Z.ai | 1 470 | 1 455–1 486 | 1 530 | MIT |
| 34 | GPT-5.5 · 5.5 | OpenAI | 1 469 | 1 463–1 476 | 11 589 | Proprietary |
| 35 | DeepSeek V4 Pro 0423 | DeepSeek | 1 468 | 1 460–1 475 | 9 800 | MIT |
| 36 | Claude Opus 4.5 | Anthropic | 1 466 | 1 460–1 472 | 12 466 | Proprietary |
| 37 | Claude Sonnet 4.5 | Anthropic | 1 465 | 1 460–1 471 | 13 935 | Proprietary |
| 38 | Claude Sonnet 4.6 | Anthropic | 1 463 | 1 457–1 470 | 11 962 | Proprietary |
| 39 | GPT-5.4 | OpenAI | 1 463 | 1 456–1 470 | 11 964 | Proprietary |
| 40 | GLM 5.2 в режиме max | Z.ai | 1 462 | 1 454–1 471 | 6 003 | MIT |
| 41 | GPT-6 Astra в режиме max | OpenAI | 1 461 | 1 431–1 492 | 373 | Proprietary |
| 42 | Claude Opus 4.5 в режиме high | Anthropic | 1 461 | 1 453–1 469 | 6 268 | Proprietary |
| 43 | Qwen3.7 Plus | Alibaba Qwen | 1 461 | 1 453–1 469 | 6 534 | Proprietary |
| 44 | Grok 4.5 | xAI | 1 459 | 1 450–1 468 | 5 131 | Proprietary |
| 45 | MiMo-V2 Pro | Xiaomi | 1 459 | 1 449–1 468 | 4 416 | Proprietary |
| 46 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 458 | 1 449–1 468 | 4 537 | Proprietary |
| 47 | GLM 5 | Z.ai | 1 457 | 1 448–1 466 | 4 417 | MIT |
| 48 | GPT-5.2 Chat | OpenAI | 1 457 | 1 449–1 465 | 6 494 | Proprietary |
| 49 | Grok 4.20 · beta-0309-reasoning | xAI | 1 457 | 1 450–1 464 | 11 430 | Proprietary |
| 50 | Qwen3.6 Max Preview | Alibaba Qwen | 1 456 | 1 436–1 476 | 903 | Proprietary |
| 51 | Gemini 3 Flash Preview в режиме minimal | 1 455 | 1 450–1 461 | 15 441 | Proprietary | |
| 52 | Claude Sonnet 5 в режиме high | Anthropic | 1 454 | 1 446–1 463 | 6 132 | Proprietary |
| 53 | ChatGPT-4o (latest) | OpenAI | 1 454 | 1 449–1 460 | 14 077 | Proprietary |
| 54 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 453 | 1 444–1 462 | 4 702 | Proprietary |
| 55 | Gemini 2.5 Pro | 1 453 | 1 448–1 457 | 20 747 | Proprietary | |
| 56 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 452 | 1 436–1 468 | 1 439 | MIT |
| 57 | Kimi K2.6 | Moonshot AI | 1 452 | 1 444–1 460 | 6 553 | Modified MIT |
| 58 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 451 | 1 446–1 457 | 13 753 | Proprietary |
| 59 | Gemma 4 31B | 1 451 | 1 433–1 469 | 1 078 | Apache 2.0 | |
| 60 | Grok 4.20 Multi-Agent | xAI | 1 451 | 1 444–1 458 | 10 021 | Proprietary |
| 61 | Grok 4.20 · beta1 | xAI | 1 449 | 1 440–1 459 | 4 366 | Proprietary |
| 62 | DeepSeek V4 Flash 0423 | DeepSeek | 1 448 | 1 441–1 456 | 8 822 | MIT |
| 63 | GPT-5.1 в режиме high | OpenAI | 1 448 | 1 440–1 455 | 6 994 | Proprietary |
| 64 | GLM 4.7 | Z.ai | 1 447 | 1 433–1 461 | 1 866 | MIT |
| 65 | Qwen3 Max · preview | Alibaba Qwen | 1 447 | 1 438–1 456 | 4 519 | Proprietary |
| 66 | Qwen3.5 397B A17B | Alibaba Qwen | 1 446 | 1 440–1 452 | 12 773 | Apache 2.0 |
| 67 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 445 | 1 438–1 452 | 8 343 | Proprietary |
| 68 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 445 | 1 437–1 452 | 9 120 | MIT |
| 69 | GPT-4.5 Preview | OpenAI | 1 444 | 1 431–1 458 | 2 052 | Proprietary |
| 70 | Kimi K2.5 · thinking | Moonshot AI | 1 444 | 1 438–1 450 | 12 465 | Modified MIT |
| 71 | MiMo-V2.5 | Xiaomi | 1 444 | 1 436–1 452 | 7 902 | MIT |
| 72 | MiMo-V2 Omni | Xiaomi | 1 443 | 1 432–1 454 | 3 293 | Proprietary |
| 73 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 443 | 1 420–1 467 | 586 | Proprietary |
| 74 | Claude Opus 4.1 · 20250805 | Anthropic | 1 442 | 1 437–1 448 | 12 900 | Proprietary |
| 75 | Gemini 3.5 Flash Lite | 1 442 | 1 433–1 452 | 4 360 | Proprietary | |
| 76 | MiniMax M3 | MiniMax | 1 442 | 1 435–1 450 | 8 605 | MiniMax Community License |
| 77 | Seed 2.0 Pro | ByteDance Seed | 1 441 | 1 435–1 447 | 13 034 | Proprietary |
| 78 | Qwen3.8 27B | Alibaba Qwen | 1 440 | 1 426–1 455 | 1 812 | Apache 2.0 |
| 79 | Qwen3.6 Plus | Alibaba Qwen | 1 440 | 1 432–1 447 | 8 019 | Proprietary |
| 80 | Gemma 4 26B A4B | 1 440 | 1 422–1 457 | 1 092 | Apache 2.0 | |
| 81 | Hy3 | Tencent | 1 440 | 1 423–1 456 | 1 349 | Apache 2.0 |
| 82 | Grok 4.1 · 4.1 | xAI | 1 437 | 1 431–1 444 | 11 983 | Proprietary |
| 83 | Inkling | Thinking Machines Lab | 1 436 | 1 426–1 445 | 4 205 | Apache 2.0 |
| 84 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 435 | 1 421–1 450 | 1 633 | Proprietary |
| 85 | GPT-5.5 · 5.5-instant | OpenAI | 1 434 | 1 425–1 443 | 4 900 | Proprietary |
| 86 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 434 | 1 425–1 443 | 4 712 | Proprietary |
| 87 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 433 | 1 428–1 438 | 16 655 | Apache 2.0 |
| 88 | ERNIE 5.0 · 0110 | Baidu | 1 433 | 1 424–1 441 | 5 650 | Proprietary |
| 89 | GPT-5.1 | OpenAI | 1 432 | 1 425–1 440 | 7 944 | Proprietary |
| 90 | GLM 5V Turbo | Z.ai | 1 431 | 1 416–1 447 | 1 557 | Proprietary |
| 91 | GPT-5.4 Mini в режиме high | OpenAI | 1 431 | 1 424–1 438 | 11 168 | Proprietary |
| 92 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 431 | 1 423–1 438 | 8 679 | MIT |
| 93 | Grok 4.1 · 4.1-thinking | xAI | 1 431 | 1 424–1 437 | 11 783 | Proprietary |
| 94 | Mistral Large 3 2512 | Mistral AI | 1 430 | 1 424–1 435 | 11 778 | Apache 2.0 |
| 95 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 428 | 1 414–1 442 | 1 920 | Apache 2.0 |
| 96 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 426 | 1 419–1 434 | 8 038 | MIT |
| 97 | GPT-5 | OpenAI | 1 426 | 1 418–1 434 | 5 674 | Proprietary |
| 98 | GLM 4.6 | Z.ai | 1 426 | 1 418–1 434 | 5 470 | MIT |
| 99 | Kimi K2.5 · instant | Moonshot AI | 1 425 | 1 410–1 441 | 1 416 | Modified MIT |
| 100 | Grok 3 | xAI | 1 425 | 1 416–1 434 | 4 650 | Proprietary |
| 101 | Hunyuan Vision 1.5 | Tencent | 1 423 | 1 395–1 451 | 426 | Proprietary |
| 102 | Inkling Small | Thinking Machines Lab | 1 423 | 1 411–1 434 | 3 067 | Apache 2.0 |
| 103 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 423 | 1 415–1 430 | 6 749 | MIT |
| 104 | ERNIE 5.0 · preview-1022 | Baidu | 1 422 | 1 401–1 444 | 696 | Proprietary |
| 105 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 422 | 1 409–1 435 | 1 938 | MIT |
| 106 | Mistral Medium 3.5 | Mistral AI | 1 422 | 1 407–1 436 | 1 801 | Modified MIT |
| 107 | Grok 4.6 в режиме high | xAI | 1 421 | 1 408–1 433 | 2 407 | Proprietary |
| 108 | GPT-5.2 в режиме high | OpenAI | 1 419 | 1 412–1 426 | 8 799 | Proprietary |
| 109 | Mistral Medium 3.1 | Mistral AI | 1 418 | 1 413–1 424 | 15 958 | Proprietary |
| 110 | GPT-5.2 | OpenAI | 1 417 | 1 412–1 423 | 15 027 | Proprietary |
| 111 | Gemini 3.1 Flash Lite Preview | 1 417 | 1 410–1 424 | 10 972 | Proprietary | |
| 112 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 417 | 1 408–1 426 | 5 019 | Proprietary |
| 113 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 416 | 1 406–1 425 | 3 988 | Apache 2.0 |
| 114 | GLM 4.5 | Z.ai | 1 416 | 1 406–1 426 | 3 720 | MIT |
| 115 | Grok 4 | xAI | 1 415 | 1 408–1 423 | 6 184 | Proprietary |
| 116 | Kimi K2 Thinking | Moonshot AI | 1 414 | 1 408–1 420 | 10 574 | Modified MIT |
| 117 | Grok 4 Fast · chat | xAI | 1 414 | 1 396–1 431 | 1 084 | Proprietary |
| 118 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 414 | 1 405–1 422 | 4 948 | Apache 2.0 |
| 119 | MiniMax M2.7 | MiniMax | 1 413 | 1 406–1 419 | 12 145 | Modified MIT |
| 120 | DeepSeek V3.1 Terminus · terminus-thinking | DeepSeek | 1 413 | 1 388–1 437 | 577 | MIT |
| 121 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 413 | 1 397–1 428 | 1 386 | MIT |
| 122 | Nemotron 3 Ultra | NVIDIA | 1 412 | 1 397–1 427 | 1 688 | OpenMDW-1.1 |
| 123 | Hy3 preview | Tencent | 1 412 | 1 395–1 430 | 1 160 | tencent-hunyuan-community |
| 124 | GPT-5.3 Chat | OpenAI | 1 412 | 1 404–1 420 | 6 164 | Proprietary |
| 125 | ERNIE 5.0 · preview-1203 | Baidu | 1 411 | 1 396–1 427 | 1 503 | Proprietary |
| 126 | Qwen3.5-27B | Alibaba Qwen | 1 410 | 1 401–1 419 | 4 613 | Apache 2.0 |
| 127 | Claude Haiku 4.5 | Anthropic | 1 409 | 1 404–1 414 | 22 779 | Proprietary |
| 128 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 409 | 1 395–1 423 | 1 824 | MIT |
| 129 | Gemini 2.5 Flash · flash | 1 408 | 1 404–1 413 | 21 242 | Proprietary | |
| 130 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 407 | 1 392–1 423 | 1 358 | Apache 2.0 |
| 131 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 407 | 1 384–1 430 | 608 | Proprietary |
| 132 | o3 | OpenAI | 1 405 | 1 398–1 412 | 9 659 | Proprietary |
| 133 | Grok 4.3 | xAI | 1 405 | 1 398–1 412 | 12 101 | Proprietary |
| 134 | R1 0528 | DeepSeek | 1 404 | 1 393–1 415 | 2 850 | MIT |
| 135 | Muse Glimmer 30B | Meta | 1 404 | 1 379–1 429 | 594 | Apache-2.0 |
| 136 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 404 | 1 397–1 411 | 8 182 | MIT |
| 137 | Step 3.5 Flash | StepFun | 1 404 | 1 397–1 410 | 9 880 | Apache 2.0 |
| 138 | LongCat-Flash-Chat · chat | Meituan | 1 403 | 1 390–1 416 | 1 969 | MIT |
| 139 | Grok 4 Fast · reasoning | xAI | 1 403 | 1 392–1 414 | 2 914 | Proprietary |
| 140 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 401 | 1 389–1 412 | 2 418 | MIT |
| 141 | GPT-5 в режиме high | OpenAI | 1 400 | 1 391–1 408 | 4 939 | Proprietary |
| 142 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 398 | 1 390–1 406 | 6 557 | Apache 2.0 |
| 143 | GPT-4.1 | OpenAI | 1 398 | 1 391–1 405 | 8 705 | Proprietary |
| 144 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 397 | 1 388–1 405 | 6 056 | Proprietary |
| 145 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 395 | 1 387–1 404 | 5 630 | Proprietary | |
| 146 | MiniMax M2.1 | MiniMax | 1 395 | 1 383–1 406 | 2 727 | MIT |
| 147 | Solar Pro 4 | Upstage | 1 393 | 1 362–1 424 | 373 | Proprietary |
| 148 | Hunyuan T1 | Tencent | 1 393 | 1 371–1 415 | 668 | Proprietary |
| 149 | Qwen3.5-Flash | Alibaba Qwen | 1 393 | 1 386–1 399 | 10 304 | Proprietary |
| 150 | R1 | DeepSeek | 1 391 | 1 379–1 403 | 2 418 | MIT |
| 151 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 389 | 1 381–1 398 | 5 209 | Apache 2.0 |
| 152 | Grok 4.1 Fast | xAI | 1 389 | 1 382–1 396 | 9 881 | Proprietary |
| 153 | DeepSeek V3 0324 | DeepSeek | 1 389 | 1 381–1 396 | 7 701 | MIT |
| 154 | Claude Opus 4 · 20250514 | Anthropic | 1 389 | 1 381–1 396 | 7 388 | Proprietary |
| 155 | DeepSeek V3.1 Terminus · terminus | DeepSeek | 1 388 | 1 366–1 411 | 662 | MIT |
| 156 | Hunyuan TurboS · 20250416 | Tencent | 1 387 | 1 373–1 401 | 1 681 | Proprietary |
| 157 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 386 | 1 369–1 403 | 1 253 | Apache 2.0 |
| 158 | Mistral Medium 3 | Mistral AI | 1 383 | 1 375–1 392 | 5 724 | Proprietary |
| 159 | GPT-5.4 Nano в режиме high | OpenAI | 1 382 | 1 375–1 389 | 11 127 | Proprietary |
| 160 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 379 | 1 370–1 388 | 4 032 | Proprietary |
| 161 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 378 | 1 355–1 402 | 584 | Proprietary |
| 162 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 378 | 1 368–1 387 | 3 949 | Apache 2.0 |
| 163 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 377 | 1 363–1 390 | 1 860 | MIT |
| 164 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 376 | 1 368–1 385 | 5 643 | Proprietary |
| 165 | GLM 4.5 Air | Z.ai | 1 370 | 1 362–1 379 | 4 973 | MIT |
| 166 | Kimi K2 0711 | Moonshot AI | 1 370 | 1 361–1 379 | 4 571 | Modified MIT |
| 167 | o1 · preview | OpenAI | 1 369 | 1 359–1 378 | 5 871 | Proprietary |
| 168 | Nemotron 3 Super | NVIDIA | 1 368 | 1 352–1 385 | 1 299 | NVIDIA Open Model |
| 169 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 367 | 1 359–1 374 | 7 953 | Proprietary | |
| 170 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 366 | 1 352–1 380 | 1 878 | Proprietary |
| 171 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 366 | 1 356–1 375 | 4 267 | Apache 2.0 |
| 172 | Claude Sonnet 4 · 20250514 | Anthropic | 1 365 | 1 357–1 373 | 6 820 | Proprietary |
| 173 | Kimi K2 0905 | Moonshot AI | 1 365 | 1 352–1 378 | 2 031 | Modified MIT |
| 174 | Qwen2.5 Max | Alibaba Qwen | 1 365 | 1 356–1 373 | 4 815 | Proprietary |
| 175 | MiniMax M2 | MiniMax | 1 365 | 1 347–1 382 | 1 088 | Apache 2.0 |
| 176 | GPT-5 Mini в режиме high | OpenAI | 1 363 | 1 354–1 372 | 4 382 | Proprietary |
| 177 | MiniMax M2.5 | MiniMax | 1 362 | 1 355–1 370 | 7 336 | Modified MIT |
| 178 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 361 | 1 351–1 370 | 4 191 | Apache 2.0 |
| 179 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 360 | 1 352–1 368 | 5 477 | Proprietary | |
| 180 | GLM 4.6V | Z.ai | 1 360 | 1 333–1 386 | 475 | MIT |
| 181 | Mercury 2 | Inception Labs | 1 357 | 1 331–1 382 | 519 | Proprietary |
| 182 | Grok 3 Mini в режиме high | xAI | 1 356 | 1 344–1 368 | 2 497 | Proprietary |
| 183 | o1 · 2024-12-17 | OpenAI | 1 355 | 1 346–1 364 | 4 399 | Proprietary |
| 184 | GPT-4.1 Mini | OpenAI | 1 354 | 1 346–1 361 | 7 018 | Proprietary |
| 185 | Amazon Nova Experimental Chat · 10-09 | Amazon | 1 353 | 1 327–1 379 | 471 | Proprietary |
| 186 | Grok 3 Mini | xAI | 1 351 | 1 341–1 361 | 3 399 | Proprietary |
| 187 | INTELLECT-3 | Prime Intellect | 1 351 | 1 330–1 371 | 767 | MIT |
| 188 | Gemini 2.0 Flash | 1 350 | 1 343–1 357 | 6 833 | Proprietary | |
| 189 | DeepSeek V3 | DeepSeek | 1 349 | 1 339–1 359 | 3 747 | DeepSeek |
| 190 | o4 Mini | OpenAI | 1 349 | 1 342–1 356 | 7 736 | Proprietary |
| 191 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 346 | 1 333–1 359 | 2 216 | Apache 2.0 |
| 192 | Granite 4.2 30B | IBM Granite | 1 346 | 1 319–1 373 | 539 | Apache 2.0 |
| 193 | Gemma 3 27B | 1 345 | 1 338–1 353 | 6 726 | Gemma | |
| 194 | GLM 4.7 Flash | Z.ai | 1 344 | 1 331–1 357 | 2 057 | MIT |
| 195 | Mistral Small 3.2 24B | Mistral AI | 1 343 | 1 333–1 354 | 3 003 | Apache 2.0 |
| 196 | Trinity Large | Arcee AI | 1 342 | 1 333–1 351 | 5 316 | Apache 2.0 |
| 197 | Step-1o Turbo | StepFun | 1 340 | 1 324–1 357 | 1 359 | Proprietary |
| 198 | Step 3 | StepFun | 1 340 | 1 322–1 358 | 1 038 | Apache 2.0 |
| 199 | gpt-oss-120b | OpenAI | 1 339 | 1 331–1 348 | 4 862 | Apache 2.0 |
| 200 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 339 | 1 332–1 347 | 6 829 | Proprietary |
| 201 | Command A | Cohere | 1 339 | 1 332–1 345 | 9 271 | CC-BY-NC-4.0 |
| 202 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 337 | 1 329–1 346 | 6 147 | Proprietary |
| 203 | Nova 2 Lite | Amazon | 1 337 | 1 324–1 350 | 2 047 | Proprietary |
| 204 | Qwen-Plus | Alibaba Qwen | 1 336 | 1 317–1 355 | 842 | Proprietary |
| 205 | MiniMax M1 | MiniMax | 1 335 | 1 327–1 343 | 5 425 | Apache 2.0 |
| 206 | Gemma 3 12B | 1 334 | 1 308–1 359 | 471 | Gemma | |
| 207 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 333 | 1 309–1 357 | 562 | Nvidia Open |
| 208 | GLM 4.5V | Z.ai | 1 332 | 1 312–1 352 | 855 | MIT |
| 209 | Nemotron 3.5 Lightning | NVIDIA | 1 332 | 1 314–1 349 | 1 256 | OpenMDW-1.1 |
| 210 | Qwen3 32B | Alibaba Qwen | 1 331 | 1 308–1 354 | 590 | Apache 2.0 |
| 211 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 326 | 1 321–1 332 | 15 908 | Proprietary |
| 212 | Ling-flash-2.0 | inclusionAI | 1 324 | 1 307–1 342 | 1 113 | MIT |
| 213 | GLM-4-Plus · plus-0111 | Z.ai | 1 324 | 1 305–1 343 | 851 | Proprietary |
| 214 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 323 | 1 311–1 335 | 2 568 | NVIDIA Open Model |
| 215 | Trinity Large Thinking | Arcee AI | 1 322 | 1 313–1 331 | 5 379 | Apache 2.0 |
| 216 | o3 Mini High | OpenAI | 1 321 | 1 309–1 332 | 2 535 | Proprietary |
| 217 | Hunyuan TurboS · 20250226 | Tencent | 1 320 | 1 287–1 353 | 293 | Proprietary |
| 218 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 1 317 | 1 287–1 347 | 374 | Nvidia Open Model |
| 219 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 315 | 1 284–1 347 | 301 | Nvidia |
| 220 | o1-mini | OpenAI | 1 314 | 1 307–1 322 | 9 326 | Proprietary |
| 221 | QwQ 32B | Alibaba Qwen | 1 313 | 1 304–1 323 | 3 756 | Apache 2.0 |
| 222 | o3 Mini | OpenAI | 1 313 | 1 306–1 320 | 9 312 | Proprietary |
| 223 | Yi-Lightning | 01.AI | 1 311 | 1 301–1 321 | 4 748 | Proprietary |
| 224 | GPT-5 Nano в режиме high | OpenAI | 1 311 | 1 295–1 327 | 1 427 | Proprietary |
| 225 | OLMo 3.1 32B Instruct | Ai2 | 1 309 | 1 296–1 323 | 2 075 | Apache 2.0 |
| 226 | Qwen3 30B A3B | Alibaba Qwen | 1 308 | 1 299–1 317 | 4 406 | Apache 2.0 |
| 227 | Gemini 2.0 Flash-Lite | 1 307 | 1 298–1 317 | 3 586 | Proprietary | |
| 228 | Hunyuan Turbo | Tencent | 1 304 | 1 275–1 333 | 323 | Proprietary |
| 229 | GPT-4o (2024-05-13) | OpenAI | 1 302 | 1 295–1 309 | 20 434 | Proprietary |
| 230 | Qwen2.5 Plus | Alibaba Qwen | 1 299 | 1 286–1 313 | 1 789 | Proprietary |
| 231 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 299 | 1 292–1 307 | 14 950 | Proprietary |
| 232 | Llama 3.1 405B Instruct · bf16 | Meta | 1 297 | 1 289–1 305 | 6 797 | Llama 3.1 Community |
| 233 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 297 | 1 281–1 313 | 1 178 | DeepSeek |
| 234 | Gemini 1.5 Pro · 1.5-pro-002 | 1 296 | 1 289–1 303 | 10 053 | Proprietary | |
| 235 | Granite 4.2 8B | IBM Granite | 1 295 | 1 265–1 324 | 500 | Apache 2.0 |
| 236 | Granite 4.2 3B | IBM Granite | 1 293 | 1 263–1 322 | 504 | Apache 2.0 |
| 237 | Step-2 16k | StepFun | 1 293 | 1 273–1 313 | 805 | Proprietary |
| 238 | OLMo 3 32B Think | Ai2 | 1 291 | 1 268–1 313 | 746 | Apache 2.0 |
| 239 | Grok 2 | xAI | 1 290 | 1 283–1 297 | 10 743 | Proprietary |
| 240 | Llama 4 Maverick | Meta | 1 289 | 1 282–1 297 | 6 689 | Llama 4 |
| 241 | GLM-4-Plus · plus | Z.ai | 1 289 | 1 280–1 299 | 4 844 | Proprietary |
| 242 | Llama 3.1 405B Instruct · fp8 | Meta | 1 288 | 1 280–1 295 | 11 136 | Llama 3.1 Community |
| 243 | Athene V2 Chat | Nexusflow | 1 286 | 1 277–1 295 | 4 304 | NexusFlow |
| 244 | Ring-flash-2.0 | inclusionAI | 1 285 | 1 268–1 303 | 1 146 | MIT |
| 245 | GPT-4o-mini (2024-07-18) | OpenAI | 1 285 | 1 279–1 292 | 12 279 | Proprietary |
| 246 | Hunyuan Large | Tencent | 1 285 | 1 259–1 310 | 510 | Proprietary |
| 247 | Mercury | Inception Labs | 1 285 | 1 252–1 317 | 329 | Proprietary |
| 248 | Llama 4 Scout | Meta | 1 280 | 1 272–1 289 | 5 245 | Llama |
| 249 | Llama 3.3 70B Instruct | Meta | 1 280 | 1 273–1 286 | 9 065 | Llama-3.3 |
| 250 | GPT-4o (2024-08-06) | OpenAI | 1 279 | 1 271–1 287 | 8 201 | Proprietary |
| 251 | Qwen Max | Alibaba Qwen | 1 277 | 1 266–1 288 | 3 105 | Qwen |
| 252 | GPT-4.1 Nano | OpenAI | 1 277 | 1 259–1 295 | 1 000 | Proprietary |
| 253 | Gemma 3n E4B | 1 276 | 1 266–1 287 | 3 297 | Gemma | |
| 254 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 276 | 1 259–1 292 | 1 227 | Llama 3.1 |
| 255 | Claude 3 Opus | Anthropic | 1 276 | 1 269–1 282 | 31 134 | Proprietary |
| 256 | Hunyuan Standard | Tencent | 1 275 | 1 251–1 299 | 570 | Proprietary |
| 257 | Magistral Medium | Mistral AI | 1 274 | 1 259–1 289 | 1 606 | Proprietary |
| 258 | Granite 4.1 8B | IBM Granite | 1 274 | 1 248–1 299 | 673 | Apache 2.0 |
| 259 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 272 | 1 264–1 280 | 7 250 | Qwen |
| 260 | Mistral Small 3.1 24B | Mistral AI | 1 270 | 1 262–1 279 | 5 427 | Apache 2.0 |
| 261 | Gemini 1.5 Pro · advanced-0514 | 1 269 | 1 260–1 279 | 7 776 | Proprietary | |
| 262 | Gemini 1.5 Pro · 1.5-pro-001 | 1 268 | 1 261–1 276 | 13 712 | Proprietary | |
| 263 | gpt-oss-20b | OpenAI | 1 268 | 1 254–1 283 | 1 785 | Apache 2.0 |
| 264 | GPT-4 Turbo | OpenAI | 1 267 | 1 260–1 275 | 15 255 | Proprietary |
| 265 | Grok 2 Mini | xAI | 1 266 | 1 258–1 273 | 9 399 | Proprietary |
| 266 | Claude 3.5 Haiku | Anthropic | 1 265 | 1 259–1 271 | 11 191 | Proprietary |
| 267 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 261 | 1 251–1 271 | 4 473 | DeepSeek |
| 268 | Mistral Large | Mistral AI | 1 260 | 1 251–1 269 | 4 438 | MRL |
| 269 | Mistral Large 2407 | Mistral AI | 1 258 | 1 250–1 265 | 8 296 | Mistral Research |
| 270 | Athene 70B | Nexusflow | 1 257 | 1 247–1 268 | 3 344 | CC-BY-NC-4.0 |
| 271 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 257 | 1 249–1 265 | 14 867 | Proprietary |
| 272 | Llama 3.1 70B Instruct | Meta | 1 256 | 1 249–1 264 | 10 405 | Llama 3.1 Community |
| 273 | Gemma 3 4B | 1 255 | 1 232–1 279 | 565 | Gemma | |
| 274 | Gemini 1.5 Flash · 002 | 1 253 | 1 245–1 261 | 6 324 | Proprietary | |
| 275 | Hunyuan Large Vision | Tencent | 1 253 | 1 233–1 273 | 942 | Proprietary |
| 276 | Llama 3.1 Tulu 3 70B | Ai2 | 1 252 | 1 229–1 275 | 537 | Llama 3.1 |
| 277 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 251 | 1 243–1 258 | 14 381 | Proprietary |
| 278 | OLMo 3.1 32B Think | Ai2 | 1 250 | 1 232–1 268 | 1 148 | Apache 2.0 |
| 279 | Nova Pro 1.0 | Amazon | 1 246 | 1 236–1 255 | 4 060 | Proprietary |
| 280 | Granite 4.0 H Small | IBM Granite | 1 244 | 1 223–1 264 | 940 | Apache 2.0 |
| 281 | Gemini 1.5 Flash · 001 | 1 228 | 1 220–1 236 | 11 065 | Proprietary | |
| 282 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 227 | 1 207–1 247 | 781 | Llama 3.1 |
| 283 | Claude 3 Sonnet | Anthropic | 1 227 | 1 219–1 235 | 17 062 | Proprietary |
| 284 | Reka Core | Reka AI | 1 227 | 1 211–1 243 | 1 255 | Proprietary |
| 285 | Gemma 2 27B | 1 224 | 1 217–1 230 | 13 588 | Gemma license | |
| 286 | Llama 3 70B Instruct | Meta | 1 223 | 1 216–1 231 | 21 421 | Llama 3 Community |
| 287 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 223 | 1 205–1 240 | 1 024 | Apache 2.0 |
| 288 | OLMo 2 32B Instruct | Ai2 | 1 221 | 1 192–1 251 | 363 | Apache-2.0 |
| 289 | Jamba 1.5 Large | AI21 Labs | 1 220 | 1 206–1 235 | 1 634 | Jamba Open |
| 290 | Gemma 2 9B IT SimPO | Princeton NLP | 1 220 | 1 205–1 234 | 1 660 | MIT |
| 291 | GLM-4 | Z.ai | 1 218 | 1 204–1 233 | 1 631 | Proprietary |
| 292 | Mistral Small 3 | Mistral AI | 1 217 | 1 204–1 230 | 2 151 | Apache 2.0 |
| 293 | Nemotron-4 340B Instruct | NVIDIA | 1 214 | 1 202–1 225 | 3 098 | NVIDIA Open Model |
| 294 | Command R+ (08-2024) | Cohere | 1 213 | 1 199–1 227 | 1 826 | CC-BY-NC-4.0 |
| 295 | GPT-4 | OpenAI | 1 206 | 1 196–1 216 | 7 992 | Proprietary |
| 296 | Phi 4 | Microsoft | 1 206 | 1 196–1 216 | 3 517 | MIT |
| 297 | Nova Lite 1.0 | Amazon | 1 199 | 1 188–1 210 | 3 166 | Proprietary |
| 298 | Qwen2 72B Instruct | Alibaba Qwen | 1 196 | 1 187–1 205 | 6 698 | Qianwen LICENSE |
| 299 | Gemma 2 9B | 1 194 | 1 186–1 201 | 9 730 | Gemma license | |
| 300 | Command R+ | Cohere | 1 192 | 1 184–1 201 | 12 101 | CC-BY-NC-4.0 |
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.