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 «экспертные запросы», 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 high | Anthropic | 1 557 | 1 547–1 567 | 4 711 | Proprietary |
| 2 | Claude Opus 4.6 в режиме high | Anthropic | 1 545 | 1 536–1 553 | 6 399 | Proprietary |
| 3 | Claude Opus 5 в режиме max | Anthropic | 1 543 | 1 529–1 556 | 2 212 | Proprietary |
| 4 | Claude Fable 5.1 в режиме max | Anthropic | 1 542 | 1 514–1 570 | 468 | Proprietary |
| 5 | Claude Opus 4.6 | Anthropic | 1 537 | 1 529–1 545 | 7 629 | Proprietary |
| 6 | Claude Fable 5 | Anthropic | 1 533 | 1 522–1 544 | 3 257 | Proprietary |
| 7 | Muse Spark 1.3 в режиме max | Meta | 1 523 | 1 497–1 549 | 512 | Proprietary |
| 8 | Claude Opus 4.7 | Anthropic | 1 520 | 1 512–1 529 | 6 434 | Proprietary |
| 9 | Claude Opus 4.7 в режиме high | Anthropic | 1 518 | 1 509–1 526 | 6 210 | Proprietary |
| 10 | GLM 5.3 в режиме max | Z.ai | 1 517 | 1 500–1 535 | 1 151 | MIT |
| 11 | Kimi K3 в режиме max | Moonshot AI | 1 515 | 1 501–1 529 | 1 863 | Kimi K3 license |
| 12 | Gemini 3.7 Flash в режиме high | 1 514 | 1 491–1 537 | 649 | Proprietary | |
| 13 | GLM 5.3 Flash | Z.ai | 1 514 | 1 496–1 533 | 1 068 | MIT |
| 14 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 514 | 1 503–1 526 | 2 954 | Proprietary |
| 15 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 509 | 1 495–1 523 | 1 847 | Proprietary |
| 16 | GPT-5.4 в режиме high | OpenAI | 1 507 | 1 498–1 516 | 5 741 | Proprietary |
| 17 | GPT-5.5 в режиме high | OpenAI | 1 506 | 1 498–1 515 | 6 732 | Proprietary |
| 18 | GPT-5.5 · 5.5 | OpenAI | 1 503 | 1 495–1 511 | 6 976 | Proprietary |
| 19 | MiMo-V2.5-Pro | Xiaomi | 1 503 | 1 494–1 511 | 5 796 | MIT |
| 20 | Claude Sonnet 4.6 | Anthropic | 1 502 | 1 494–1 510 | 6 684 | Proprietary |
| 21 | Gemini 3.8 Flash в режиме high | 1 501 | 1 477–1 525 | 586 | Proprietary | |
| 22 | Claude Opus 4.8 в режиме high | Anthropic | 1 500 | 1 491–1 509 | 5 635 | Proprietary |
| 23 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 499 | 1 465–1 532 | 279 | Proprietary |
| 24 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 492 | 1 481–1 503 | 3 169 | Proprietary |
| 25 | Claude Opus 4.8 | Anthropic | 1 492 | 1 484–1 501 | 5 862 | Proprietary |
| 26 | Gemini 3.6 Flash в режиме high | 1 492 | 1 481–1 503 | 3 029 | Proprietary | |
| 27 | Qwen3.8 27B | Alibaba Qwen | 1 491 | 1 474–1 509 | 1 111 | Apache 2.0 |
| 28 | Gemini 3.5 Flash в режиме high | 1 491 | 1 482–1 501 | 4 263 | Proprietary | |
| 29 | ERNIE 5.1 | Baidu | 1 491 | 1 480–1 501 | 3 414 | Proprietary |
| 30 | Claude Sonnet 5 в режиме high | Anthropic | 1 490 | 1 480–1 500 | 3 850 | Proprietary |
| 31 | Kimi K2.6 | Moonshot AI | 1 489 | 1 479–1 500 | 3 505 | Modified MIT |
| 32 | Qwen3.5 Max | Alibaba Qwen | 1 489 | 1 476–1 502 | 2 062 | Proprietary |
| 33 | Qwen3.7 Max | Alibaba Qwen | 1 488 | 1 456–1 520 | 348 | Proprietary |
| 34 | Claude Opus 4.5 | Anthropic | 1 488 | 1 480–1 497 | 5 255 | Proprietary |
| 35 | Muse Spark 1.2 в режиме xhigh | Meta | 1 488 | 1 457–1 519 | 348 | Proprietary |
| 36 | Gemini 3.5 Flash в режиме medium | 1 485 | 1 475–1 495 | 4 146 | Proprietary | |
| 37 | Gemini 3.1 Pro Preview | 1 484 | 1 477–1 490 | 10 668 | Proprietary | |
| 38 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 483 | 1 475–1 492 | 5 653 | Proprietary |
| 39 | GPT-5.4 | OpenAI | 1 480 | 1 472–1 489 | 6 098 | Proprietary |
| 40 | Muse Spark 1.1 | Meta | 1 479 | 1 468–1 490 | 3 117 | Proprietary |
| 41 | Gemini 3 Pro | 1 478 | 1 467–1 490 | 2 634 | Proprietary | |
| 42 | Claude Opus 4.5 в режиме high | Anthropic | 1 478 | 1 465–1 490 | 2 282 | Proprietary |
| 43 | GLM 5.1 | Z.ai | 1 477 | 1 468–1 486 | 5 065 | MIT |
| 44 | MiMo-V2 Pro | Xiaomi | 1 477 | 1 464–1 490 | 2 087 | Proprietary |
| 45 | GLM 5.2 в режиме max | Z.ai | 1 476 | 1 466–1 486 | 3 869 | MIT |
| 46 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 476 | 1 465–1 487 | 3 152 | Proprietary |
| 47 | Qwen3.6 Max Preview | Alibaba Qwen | 1 475 | 1 449–1 501 | 520 | Proprietary |
| 48 | Claude Sonnet 4.5 | Anthropic | 1 472 | 1 464–1 480 | 5 771 | Proprietary |
| 49 | GPT-5.1 в режиме high | OpenAI | 1 471 | 1 458–1 483 | 2 419 | Proprietary |
| 50 | Nemotron 3 Ultra | NVIDIA | 1 470 | 1 453–1 488 | 1 144 | OpenMDW-1.1 |
| 51 | Qwen3.7 Plus | Alibaba Qwen | 1 469 | 1 459–1 478 | 4 330 | Proprietary |
| 52 | Grok 4.5 | xAI | 1 468 | 1 458–1 479 | 3 399 | Proprietary |
| 53 | Kimi K2.5 · thinking | Moonshot AI | 1 466 | 1 458–1 474 | 6 425 | Modified MIT |
| 54 | Qwen3.5 397B A17B | Alibaba Qwen | 1 465 | 1 457–1 472 | 7 981 | Apache 2.0 |
| 55 | DeepSeek V4 Pro 0423 | DeepSeek | 1 465 | 1 456–1 473 | 5 440 | MIT |
| 56 | Inkling | Thinking Machines Lab | 1 464 | 1 453–1 476 | 2 971 | Apache 2.0 |
| 57 | MiniMax M3 | MiniMax | 1 463 | 1 454–1 472 | 5 154 | MiniMax Community License |
| 58 | MiMo-V2.5 | Xiaomi | 1 462 | 1 453–1 472 | 4 395 | MIT |
| 59 | Gemini 3 Flash Preview | 1 461 | 1 448–1 474 | 2 014 | Proprietary | |
| 60 | Grok 4.6 в режиме high | xAI | 1 461 | 1 447–1 475 | 1 758 | Proprietary |
| 61 | Qwen3 Max · preview | Alibaba Qwen | 1 459 | 1 443–1 476 | 1 291 | Proprietary |
| 62 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 459 | 1 431–1 488 | 416 | Apache 2.0 |
| 63 | Muse Spark | Meta | 1 456 | 1 440–1 473 | 1 293 | Proprietary |
| 64 | Qwen3.6 Plus | Alibaba Qwen | 1 456 | 1 447–1 466 | 4 491 | Proprietary |
| 65 | Hy3 | Tencent | 1 456 | 1 435–1 476 | 838 | Apache 2.0 |
| 66 | GLM 5 | Z.ai | 1 455 | 1 443–1 468 | 2 390 | MIT |
| 67 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 454 | 1 435–1 474 | 916 | MIT |
| 68 | Gemini 2.5 Pro | 1 454 | 1 447–1 461 | 7 917 | Proprietary | |
| 69 | Gemma 4 31B | 1 453 | 1 427–1 479 | 467 | Apache 2.0 | |
| 70 | GLM 5V Turbo | Z.ai | 1 452 | 1 434–1 470 | 1 033 | Proprietary |
| 71 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 452 | 1 440–1 464 | 2 600 | Proprietary |
| 72 | Gemma 4 26B A4B | 1 449 | 1 422–1 475 | 421 | Apache 2.0 | |
| 73 | MiMo-V2 Omni | Xiaomi | 1 448 | 1 435–1 462 | 1 967 | Proprietary |
| 74 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 448 | 1 439–1 457 | 5 054 | MIT |
| 75 | GPT-5.2 в режиме high | OpenAI | 1 446 | 1 436–1 456 | 3 741 | Proprietary |
| 76 | GPT-6 Astra в режиме max | OpenAI | 1 445 | 1 407–1 484 | 247 | Proprietary |
| 77 | LongCat-Flash-Chat · chat | Meituan | 1 444 | 1 418–1 470 | 518 | MIT |
| 78 | Inkling Small | Thinking Machines Lab | 1 444 | 1 431–1 457 | 2 168 | Apache 2.0 |
| 79 | Claude Haiku 4.5 | Anthropic | 1 444 | 1 438–1 450 | 11 259 | Proprietary |
| 80 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 444 | 1 436–1 451 | 6 102 | Apache 2.0 |
| 81 | Grok 4.20 Multi-Agent | xAI | 1 443 | 1 435–1 452 | 6 031 | Proprietary |
| 82 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 443 | 1 408–1 478 | 265 | Proprietary |
| 83 | DeepSeek V4 Flash 0423 | DeepSeek | 1 443 | 1 433–1 452 | 4 896 | MIT |
| 84 | GPT-5.1 | OpenAI | 1 442 | 1 431–1 453 | 2 837 | Proprietary |
| 85 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 441 | 1 429–1 453 | 2 439 | Proprietary |
| 86 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 440 | 1 415–1 465 | 563 | Apache 2.0 |
| 87 | Kimi K2 Thinking | Moonshot AI | 1 439 | 1 430–1 448 | 4 430 | Modified MIT |
| 88 | Kimi K2.5 · instant | Moonshot AI | 1 439 | 1 415–1 463 | 575 | Modified MIT |
| 89 | MiniMax M2.7 | MiniMax | 1 439 | 1 431–1 447 | 6 619 | Modified MIT |
| 90 | Seed 2.0 Pro | ByteDance Seed | 1 439 | 1 431–1 446 | 7 161 | Proprietary |
| 91 | Grok 4.20 · beta-0309-reasoning | xAI | 1 438 | 1 430–1 447 | 6 095 | Proprietary |
| 92 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 437 | 1 428–1 447 | 4 899 | MIT |
| 93 | GPT-5.2 Chat | OpenAI | 1 437 | 1 426–1 448 | 2 948 | Proprietary |
| 94 | MiniMax M2.1 | MiniMax | 1 436 | 1 418–1 453 | 1 088 | MIT |
| 95 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 435 | 1 423–1 447 | 2 528 | Apache 2.0 |
| 96 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 435 | 1 424–1 445 | 3 191 | MIT |
| 97 | GPT-5.4 Mini в режиме high | OpenAI | 1 435 | 1 426–1 443 | 5 775 | Proprietary |
| 98 | Mistral Medium 3.5 | Mistral AI | 1 434 | 1 417–1 452 | 1 106 | Modified MIT |
| 99 | GLM 4.6 | Z.ai | 1 434 | 1 421–1 447 | 1 949 | MIT |
| 100 | Gemini 3.5 Flash Lite | 1 433 | 1 422–1 445 | 2 946 | Proprietary | |
| 101 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 432 | 1 418–1 447 | 1 604 | Proprietary |
| 102 | GPT-5.2 | OpenAI | 1 431 | 1 423–1 439 | 7 244 | Proprietary |
| 103 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 430 | 1 419–1 442 | 2 682 | MIT |
| 104 | GLM 4.5 | Z.ai | 1 429 | 1 412–1 447 | 1 129 | MIT |
| 105 | Grok 4.20 · beta1 | xAI | 1 428 | 1 416–1 441 | 2 307 | Proprietary |
| 106 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 427 | 1 417–1 437 | 3 726 | MIT |
| 107 | Qwen3.5-27B | Alibaba Qwen | 1 425 | 1 413–1 437 | 2 440 | Apache 2.0 |
| 108 | Gemini 2.5 Flash · flash | 1 425 | 1 418–1 432 | 7 988 | Proprietary | |
| 109 | Claude Opus 4.1 · 20250805 | Anthropic | 1 425 | 1 416–1 435 | 4 041 | Proprietary |
| 110 | Hy3 preview | Tencent | 1 425 | 1 402–1 447 | 671 | tencent-hunyuan-community |
| 111 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 424 | 1 394–1 455 | 378 | Apache 2.0 |
| 112 | ERNIE 5.0 · 0110 | Baidu | 1 424 | 1 413–1 435 | 2 753 | Proprietary |
| 113 | Step 3.5 Flash | StepFun | 1 423 | 1 414–1 432 | 4 948 | Apache 2.0 |
| 114 | ERNIE 5.0 · preview-1203 | Baidu | 1 422 | 1 400–1 444 | 694 | Proprietary |
| 115 | ERNIE 5.0 · preview-1022 | Baidu | 1 421 | 1 387–1 455 | 278 | Proprietary |
| 116 | GPT-5 в режиме high | OpenAI | 1 420 | 1 405–1 435 | 1 639 | Proprietary |
| 117 | Grok 3 | xAI | 1 420 | 1 405–1 435 | 1 551 | Proprietary |
| 118 | GLM 4.7 | Z.ai | 1 419 | 1 398–1 440 | 747 | MIT |
| 119 | Gemini 3 Flash Preview в режиме minimal | 1 419 | 1 411–1 426 | 7 818 | Proprietary | |
| 120 | Grok 4.1 · 4.1-thinking | xAI | 1 419 | 1 410–1 427 | 4 658 | Proprietary |
| 121 | Mistral Large 3 2512 | Mistral AI | 1 418 | 1 410–1 426 | 6 136 | Apache 2.0 |
| 122 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 418 | 1 389–1 447 | 396 | MIT |
| 123 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 418 | 1 404–1 432 | 1 699 | Proprietary | |
| 124 | Grok 4 | xAI | 1 417 | 1 404–1 431 | 2 023 | Proprietary |
| 125 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 414 | 1 396–1 432 | 1 046 | Apache 2.0 |
| 126 | GPT-5.5 · 5.5-instant | OpenAI | 1 413 | 1 400–1 425 | 2 542 | Proprietary |
| 127 | Grok 4.1 · 4.1 | xAI | 1 413 | 1 404–1 421 | 4 760 | Proprietary |
| 128 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 412 | 1 377–1 447 | 254 | Proprietary |
| 129 | Qwen3.5-Flash | Alibaba Qwen | 1 409 | 1 401–1 418 | 5 404 | Proprietary |
| 130 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 407 | 1 396–1 419 | 2 600 | Apache 2.0 |
| 131 | Mistral Medium 3.1 | Mistral AI | 1 407 | 1 400–1 415 | 6 179 | Proprietary |
| 132 | Grok 4 Fast · reasoning | xAI | 1 407 | 1 387–1 426 | 875 | Proprietary |
| 133 | GPT-5 | OpenAI | 1 404 | 1 389–1 419 | 1 493 | Proprietary |
| 134 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 403 | 1 382–1 425 | 735 | MIT |
| 135 | Granite 4.2 30B | IBM Granite | 1 403 | 1 372–1 434 | 336 | Apache 2.0 |
| 136 | o3 | OpenAI | 1 403 | 1 391–1 414 | 2 975 | Proprietary |
| 137 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 403 | 1 378–1 427 | 535 | MIT |
| 138 | ChatGPT-4o (latest) | OpenAI | 1 400 | 1 391–1 409 | 4 412 | Proprietary |
| 139 | Solar Pro 4 | Upstage | 1 400 | 1 363–1 437 | 257 | Proprietary |
| 140 | Grok 4 Fast · chat | xAI | 1 399 | 1 365–1 433 | 291 | Proprietary |
| 141 | Grok 4.1 Fast | xAI | 1 399 | 1 389–1 408 | 4 171 | Proprietary |
| 142 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 398 | 1 376–1 421 | 644 | MIT |
| 143 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 398 | 1 377–1 419 | 708 | MIT |
| 144 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 398 | 1 371–1 424 | 450 | Proprietary |
| 145 | Gemini 3.1 Flash Lite Preview | 1 397 | 1 389–1 406 | 5 757 | Proprietary | |
| 146 | Nemotron 3 Super | NVIDIA | 1 397 | 1 374–1 420 | 603 | NVIDIA Open Model |
| 147 | GPT-5.3 Chat | OpenAI | 1 396 | 1 384–1 408 | 2 809 | Proprietary |
| 148 | Hunyuan T1 | Tencent | 1 396 | 1 358–1 434 | 217 | Proprietary |
| 149 | GPT-5.4 Nano в режиме high | OpenAI | 1 395 | 1 386–1 403 | 5 824 | Proprietary |
| 150 | GPT-4.5 Preview | OpenAI | 1 394 | 1 371–1 417 | 608 | Proprietary |
| 151 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 394 | 1 377–1 411 | 1 131 | Apache 2.0 |
| 152 | Grok 3 Mini в режиме high | xAI | 1 393 | 1 374–1 412 | 903 | Proprietary |
| 153 | R1 0528 | DeepSeek | 1 393 | 1 373–1 412 | 881 | MIT |
| 154 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 386 | 1 364–1 408 | 693 | Proprietary |
| 155 | Muse Glimmer 30B | Meta | 1 385 | 1 355–1 415 | 415 | Apache-2.0 |
| 156 | Grok 4.3 | xAI | 1 384 | 1 376–1 393 | 6 901 | Proprietary |
| 157 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 383 | 1 369–1 398 | 1 731 | Proprietary |
| 158 | GLM 4.6V | Z.ai | 1 379 | 1 336–1 421 | 180 | MIT |
| 159 | GPT-5 Mini в режиме high | OpenAI | 1 377 | 1 359–1 395 | 1 176 | Proprietary |
| 160 | Granite 4.2 8B | IBM Granite | 1 376 | 1 342–1 411 | 287 | Apache 2.0 |
| 161 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 376 | 1 352–1 400 | 619 | Apache 2.0 |
| 162 | MiniMax M2.5 | MiniMax | 1 374 | 1 363–1 384 | 3 553 | Modified MIT |
| 163 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 373 | 1 362–1 384 | 2 655 | Proprietary | |
| 164 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 373 | 1 358–1 388 | 1 662 | Proprietary |
| 165 | GLM 4.5 Air | Z.ai | 1 371 | 1 355–1 387 | 1 409 | MIT |
| 166 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 369 | 1 356–1 383 | 1 899 | Apache 2.0 |
| 167 | Claude Opus 4 · 20250514 | Anthropic | 1 369 | 1 356–1 381 | 2 248 | Proprietary |
| 168 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 366 | 1 352–1 381 | 1 597 | Proprietary | |
| 169 | Nova 2 Lite | Amazon | 1 366 | 1 344–1 387 | 744 | Proprietary |
| 170 | o3 Mini High | OpenAI | 1 364 | 1 344–1 384 | 847 | Proprietary |
| 171 | GPT-4.1 | OpenAI | 1 364 | 1 352–1 376 | 2 570 | Proprietary |
| 172 | Grok 3 Mini | xAI | 1 363 | 1 346–1 381 | 1 124 | Proprietary |
| 173 | Kimi K2 0905 | Moonshot AI | 1 363 | 1 339–1 387 | 575 | Modified MIT |
| 174 | Qwen3 32B | Alibaba Qwen | 1 362 | 1 325–1 398 | 236 | Apache 2.0 |
| 175 | Nemotron 3.5 Lightning | NVIDIA | 1 362 | 1 342–1 382 | 898 | OpenMDW-1.1 |
| 176 | GLM 4.7 Flash | Z.ai | 1 361 | 1 342–1 381 | 806 | MIT |
| 177 | o1 · 2024-12-17 | OpenAI | 1 361 | 1 344–1 378 | 1 330 | Proprietary |
| 178 | Trinity Large Thinking | Arcee AI | 1 361 | 1 349–1 373 | 2 681 | Apache 2.0 |
| 179 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 360 | 1 341–1 379 | 976 | NVIDIA Open Model |
| 180 | Ling-flash-2.0 | inclusionAI | 1 359 | 1 329–1 390 | 334 | MIT |
| 181 | Mercury 2 | Inception Labs | 1 356 | 1 319–1 393 | 233 | Proprietary |
| 182 | GLM 4.5V | Z.ai | 1 355 | 1 314–1 396 | 193 | MIT |
| 183 | gpt-oss-120b | OpenAI | 1 353 | 1 337–1 370 | 1 341 | Apache 2.0 |
| 184 | DeepSeek V3 0324 | DeepSeek | 1 353 | 1 341–1 365 | 2 301 | MIT |
| 185 | Trinity Large | Arcee AI | 1 351 | 1 339–1 363 | 2 669 | Apache 2.0 |
| 186 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 349 | 1 333–1 364 | 1 330 | Apache 2.0 |
| 187 | Kimi K2 0711 | Moonshot AI | 1 346 | 1 331–1 361 | 1 481 | Modified MIT |
| 188 | o4 Mini | OpenAI | 1 344 | 1 331–1 356 | 2 282 | Proprietary |
| 189 | Mistral Medium 3 | Mistral AI | 1 342 | 1 328–1 356 | 1 780 | Proprietary |
| 190 | Gemini 2.0 Flash | 1 339 | 1 327–1 352 | 2 195 | Proprietary | |
| 191 | Hunyuan TurboS | Tencent | 1 338 | 1 314–1 362 | 582 | Proprietary |
| 192 | Ring-flash-2.0 | inclusionAI | 1 338 | 1 305–1 370 | 327 | MIT |
| 193 | o1 · preview | OpenAI | 1 338 | 1 323–1 352 | 1 982 | Proprietary |
| 194 | GPT-4.1 Mini | OpenAI | 1 337 | 1 324–1 350 | 2 015 | Proprietary |
| 195 | R1 | DeepSeek | 1 337 | 1 318–1 357 | 848 | MIT |
| 196 | Qwen2.5 Max | Alibaba Qwen | 1 337 | 1 323–1 350 | 1 680 | Proprietary |
| 197 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 336 | 1 320–1 353 | 1 333 | Apache 2.0 |
| 198 | Claude Sonnet 4 · 20250514 | Anthropic | 1 334 | 1 321–1 348 | 2 041 | Proprietary |
| 199 | Step 3 | StepFun | 1 332 | 1 297–1 367 | 262 | Apache 2.0 |
| 200 | o3 Mini | OpenAI | 1 330 | 1 319–1 341 | 2 871 | Proprietary |
| 201 | MiniMax M2 | MiniMax | 1 330 | 1 296–1 363 | 277 | Apache 2.0 |
| 202 | Qwen-Plus | Alibaba Qwen | 1 328 | 1 298–1 357 | 358 | Proprietary |
| 203 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 327 | 1 286–1 368 | 178 | Nvidia Open |
| 204 | QwQ 32B | Alibaba Qwen | 1 323 | 1 306–1 339 | 1 209 | Apache 2.0 |
| 205 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 321 | 1 307–1 335 | 1 878 | Proprietary |
| 206 | GPT-5 Nano в режиме high | OpenAI | 1 320 | 1 288–1 352 | 336 | Proprietary |
| 207 | INTELLECT-3 | Prime Intellect | 1 318 | 1 286–1 351 | 335 | MIT |
| 208 | o1-mini | OpenAI | 1 316 | 1 304–1 327 | 3 191 | Proprietary |
| 209 | Qwen3 30B A3B | Alibaba Qwen | 1 314 | 1 298–1 330 | 1 333 | Apache 2.0 |
| 210 | MiniMax M1 | MiniMax | 1 313 | 1 299–1 328 | 1 586 | Apache 2.0 |
| 211 | Granite 4.1 8B | IBM Granite | 1 311 | 1 279–1 343 | 397 | Apache 2.0 |
| 212 | Step-1o Turbo | StepFun | 1 308 | 1 281–1 334 | 478 | Proprietary |
| 213 | Granite 4.2 3B | IBM Granite | 1 306 | 1 268–1 343 | 288 | Apache 2.0 |
| 214 | OLMo 3.1 32B Instruct | Ai2 | 1 306 | 1 283–1 328 | 721 | Apache 2.0 |
| 215 | DeepSeek V3 | DeepSeek | 1 305 | 1 289–1 322 | 1 236 | DeepSeek |
| 216 | Gemini 2.0 Flash-Lite | 1 304 | 1 288–1 321 | 1 237 | Proprietary | |
| 217 | Gemma 3 27B | 1 302 | 1 289–1 315 | 2 224 | Gemma | |
| 218 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 300 | 1 287–1 313 | 2 129 | Proprietary |
| 219 | OLMo 3.1 32B Think | Ai2 | 1 294 | 1 268–1 320 | 528 | Apache 2.0 |
| 220 | Mistral Small 3.2 24B | Mistral AI | 1 294 | 1 274–1 313 | 840 | Apache 2.0 |
| 221 | Command A | Cohere | 1 291 | 1 280–1 303 | 2 827 | CC-BY-NC-4.0 |
| 222 | Qwen2.5 Plus | Alibaba Qwen | 1 289 | 1 268–1 310 | 664 | Proprietary |
| 223 | Yi-Lightning | 01.AI | 1 285 | 1 271–1 300 | 1 533 | Proprietary |
| 224 | GLM-4-Plus · plus-0111 | Z.ai | 1 279 | 1 250–1 308 | 354 | Proprietary |
| 225 | Step-2 16k | StepFun | 1 278 | 1 248–1 309 | 310 | Proprietary |
| 226 | Gemini 1.5 Pro · 1.5-pro-002 | 1 278 | 1 267–1 289 | 3 319 | Proprietary | |
| 227 | OLMo 3 32B Think | Ai2 | 1 277 | 1 241–1 313 | 299 | Apache 2.0 |
| 228 | Hunyuan Large | Tencent | 1 275 | 1 240–1 311 | 228 | Proprietary |
| 229 | GPT-4.1 Nano | OpenAI | 1 271 | 1 241–1 301 | 328 | Proprietary |
| 230 | Athene V2 Chat | Nexusflow | 1 268 | 1 253–1 283 | 1 469 | NexusFlow |
| 231 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 266 | 1 240–1 292 | 441 | DeepSeek |
| 232 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 265 | 1 255–1 274 | 5 034 | Proprietary |
| 233 | Llama 4 Maverick | Meta | 1 259 | 1 246–1 273 | 2 010 | Llama 4 |
| 234 | Mistral Small 3.1 24B | Mistral AI | 1 258 | 1 243–1 273 | 1 650 | Apache 2.0 |
| 235 | gpt-oss-20b | OpenAI | 1 257 | 1 230–1 284 | 505 | Apache 2.0 |
| 236 | Grok 2 | xAI | 1 254 | 1 243–1 264 | 3 541 | Proprietary |
| 237 | GLM-4-Plus · plus | Z.ai | 1 252 | 1 237–1 266 | 1 608 | Proprietary |
| 238 | Hunyuan Large Vision | Tencent | 1 250 | 1 216–1 284 | 292 | Proprietary |
| 239 | GPT-4o (2024-05-13) | OpenAI | 1 250 | 1 240–1 260 | 5 887 | Proprietary |
| 240 | Qwen Max | Alibaba Qwen | 1 248 | 1 230–1 266 | 1 011 | Qwen |
| 241 | Hunyuan Standard | Tencent | 1 248 | 1 209–1 286 | 207 | Proprietary |
| 242 | Gemma 3 12B | 1 247 | 1 207–1 288 | 186 | Gemma | |
| 243 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 247 | 1 236–1 258 | 4 314 | Proprietary |
| 244 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 245 | 1 233–1 257 | 2 397 | Qwen |
| 245 | Gemma 3n E4B | 1 245 | 1 227–1 263 | 1 071 | Gemma | |
| 246 | Gemini 1.5 Pro · 1.5-pro-001 | 1 245 | 1 233–1 257 | 3 896 | Proprietary | |
| 247 | Llama 3.1 405B Instruct · fp8 | Meta | 1 243 | 1 231–1 254 | 3 123 | Llama 3.1 Community |
| 248 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 242 | 1 215–1 269 | 453 | Llama 3.1 |
| 249 | GPT-4o (2024-08-06) | OpenAI | 1 240 | 1 228–1 253 | 2 349 | Proprietary |
| 250 | Granite 4.0 H Small | IBM Granite | 1 240 | 1 202–1 278 | 270 | Apache 2.0 |
| 251 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 240 | 1 225–1 254 | 1 523 | DeepSeek |
| 252 | Grok 2 Mini | xAI | 1 238 | 1 226–1 249 | 2 759 | Proprietary |
| 253 | Llama 4 Scout | Meta | 1 236 | 1 221–1 252 | 1 524 | Llama |
| 254 | GPT-4o-mini (2024-07-18) | OpenAI | 1 234 | 1 224–1 245 | 3 549 | Proprietary |
| 255 | Gemini 1.5 Flash · 002 | 1 233 | 1 220–1 246 | 2 116 | Proprietary | |
| 256 | Mistral Large 2407 | Mistral AI | 1 232 | 1 219–1 245 | 2 505 | Mistral Research |
| 257 | Llama 3.1 405B Instruct · bf16 | Meta | 1 229 | 1 216–1 242 | 2 128 | Llama 3.1 Community |
| 258 | Athene 70B | Nexusflow | 1 228 | 1 208–1 247 | 867 | CC-BY-NC-4.0 |
| 259 | Llama 3.3 70B Instruct | Meta | 1 225 | 1 214–1 236 | 2 909 | Llama-3.3 |
| 260 | GPT-4 Turbo | OpenAI | 1 223 | 1 212–1 234 | 5 195 | Proprietary |
| 261 | Gemma 3 4B | 1 223 | 1 182–1 264 | 208 | Gemma | |
| 262 | Magistral Medium | Mistral AI | 1 223 | 1 196–1 249 | 564 | Proprietary |
| 263 | Claude 3 Opus | Anthropic | 1 222 | 1 213–1 232 | 10 374 | Proprietary |
| 264 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 220 | 1 188–1 253 | 267 | Apache 2.0 |
| 265 | Jamba 1.5 Large | AI21 Labs | 1 217 | 1 189–1 246 | 331 | Jamba Open |
| 266 | Gemini 1.5 Pro · advanced-0514 | 1 215 | 1 201–1 229 | 2 449 | Proprietary | |
| 267 | Reka Core | Reka AI | 1 213 | 1 189–1 238 | 458 | Proprietary |
| 268 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 211 | 1 199–1 223 | 4 240 | Proprietary |
| 269 | Nova Pro 1.0 | Amazon | 1 210 | 1 195–1 226 | 1 387 | Proprietary |
| 270 | Llama 3.1 70B Instruct | Meta | 1 209 | 1 197–1 220 | 2 924 | Llama 3.1 Community |
| 271 | Claude 3.5 Haiku | Anthropic | 1 208 | 1 197–1 218 | 3 515 | Proprietary |
| 272 | Mistral Large | Mistral AI | 1 208 | 1 193–1 223 | 1 510 | MRL |
| 273 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 205 | 1 193–1 217 | 4 454 | Proprietary |
| 274 | Phi 4 | Microsoft | 1 203 | 1 185–1 220 | 1 124 | MIT |
| 275 | Mistral Small 3 | Mistral AI | 1 202 | 1 181–1 222 | 754 | Apache 2.0 |
| 276 | Nova Lite 1.0 | Amazon | 1 201 | 1 184–1 218 | 1 138 | Proprietary |
| 277 | Gemini 1.5 Flash · 001 | 1 199 | 1 187–1 212 | 3 194 | Proprietary | |
| 278 | Gemini 1.5 Flash-8B | 1 185 | 1 172–1 198 | 2 111 | Proprietary | |
| 279 | Nova Micro 1.0 | Amazon | 1 183 | 1 166–1 201 | 1 094 | Proprietary |
| 280 | Aya Expanse 32B | Cohere | 1 182 | 1 168–1 196 | 1 764 | CC-BY-NC-4.0 |
| 281 | Reka Flash (2024-09) | Reka AI | 1 182 | 1 159–1 205 | 493 | Proprietary |
| 282 | DeepSeek Coder V2 | DeepSeek | 1 181 | 1 160–1 203 | 769 | DeepSeek License |
| 283 | GLM-4 | Z.ai | 1 179 | 1 154–1 204 | 514 | Proprietary |
| 284 | Command R+ (08-2024) | Cohere | 1 173 | 1 149–1 198 | 524 | CC-BY-NC-4.0 |
| 285 | Claude 3 Sonnet | Anthropic | 1 172 | 1 160–1 184 | 5 614 | Proprietary |
| 286 | Gemma 2 27B | 1 171 | 1 161–1 182 | 4 025 | Gemma license | |
| 287 | Nemotron-4 340B Instruct | NVIDIA | 1 171 | 1 152–1 190 | 1 027 | NVIDIA Open Model |
| 288 | Qwen2 72B Instruct | Alibaba Qwen | 1 171 | 1 156–1 186 | 1 763 | Qianwen LICENSE |
| 289 | Ministral 8B (2410) | Mistral AI | 1 170 | 1 140–1 199 | 332 | MRL |
| 290 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 167 | 1 134–1 200 | 265 | Llama 3.1 |
| 291 | Gemma 2 9B IT SimPO | Princeton NLP | 1 154 | 1 125–1 184 | 369 | MIT |
| 292 | Command R+ | Cohere | 1 153 | 1 141–1 166 | 4 031 | CC-BY-NC-4.0 |
| 293 | Aya Expanse 8B | Cohere | 1 152 | 1 129–1 176 | 600 | CC-BY-NC-4.0 |
| 294 | InternLM2.5 20B Chat | InternLM | 1 150 | 1 128–1 172 | 604 | Other |
| 295 | Llama 3 70B Instruct | Meta | 1 149 | 1 138–1 160 | 7 958 | Llama 3 Community |
| 296 | GPT-4 | OpenAI | 1 149 | 1 133–1 164 | 2 160 | Proprietary |
| 297 | Claude 3 Haiku | Anthropic | 1 148 | 1 137–1 159 | 6 336 | Proprietary |
| 298 | Gemma 2 9B | 1 147 | 1 135–1 159 | 2 847 | Gemma license | |
| 299 | Qwen1.5 110B Chat | Alibaba Qwen | 1 144 | 1 128–1 161 | 1 411 | Qianwen LICENSE |
| 300 | Yi-1.5 34B Chat | 01.AI | 1 144 | 1 125–1 162 | 1 049 | 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.