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: математика», 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 | 1 546 | 1 532–1 559 | 2 266 | Proprietary |
| 2 | Claude Opus 5 в режиме max | Anthropic | 1 529 | 1 511–1 548 | 1 116 | Proprietary |
| 3 | Claude Opus 4.6 в режиме high | Anthropic | 1 527 | 1 517–1 538 | 3 614 | Proprietary |
| 4 | Gemini 3.8 Flash в режиме high | 1 521 | 1 484–1 558 | 267 | Proprietary | |
| 5 | Claude Fable 5.1 в режиме max | Anthropic | 1 519 | 1 485–1 554 | 299 | Proprietary |
| 6 | Claude Opus 4.6 | Anthropic | 1 519 | 1 509–1 529 | 4 136 | Proprietary |
| 7 | Gemini 3.7 Flash в режиме high | 1 516 | 1 484–1 547 | 353 | Proprietary | |
| 8 | Claude Fable 5 | Anthropic | 1 512 | 1 497–1 527 | 1 651 | Proprietary |
| 9 | GLM 5.3 в режиме max | Z.ai | 1 509 | 1 485–1 533 | 583 | MIT |
| 10 | GLM 5.3 Flash | Z.ai | 1 508 | 1 483–1 533 | 569 | MIT |
| 11 | Claude Opus 4.7 в режиме high | Anthropic | 1 507 | 1 496–1 519 | 3 319 | Proprietary |
| 12 | Kimi K3 в режиме max | Moonshot AI | 1 499 | 1 481–1 518 | 1 054 | Kimi K3 license |
| 13 | Gemini 3.6 Flash в режиме high | 1 499 | 1 483–1 514 | 1 485 | Proprietary | |
| 14 | Claude Opus 4.7 | Anthropic | 1 497 | 1 486–1 508 | 3 476 | Proprietary |
| 15 | GPT-5.4 в режиме high | OpenAI | 1 496 | 1 485–1 508 | 3 236 | Proprietary |
| 16 | Claude Opus 4.8 в режиме high | Anthropic | 1 495 | 1 484–1 507 | 2 764 | Proprietary |
| 17 | GPT-5.5 в режиме high | OpenAI | 1 494 | 1 483–1 505 | 3 640 | Proprietary |
| 18 | Qwen3.5 Max | Alibaba Qwen | 1 493 | 1 476–1 510 | 1 226 | Proprietary |
| 19 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 492 | 1 476–1 508 | 1 482 | Proprietary |
| 20 | Muse Spark 1.3 в режиме max | Meta | 1 491 | 1 455–1 528 | 257 | Proprietary |
| 21 | MiMo-V2.5-Pro | Xiaomi | 1 491 | 1 479–1 502 | 3 215 | MIT |
| 22 | Muse Spark 1.1 | Meta | 1 490 | 1 474–1 505 | 1 482 | Proprietary |
| 23 | Gemini 3.5 Flash в режиме medium | 1 489 | 1 475–1 503 | 1 958 | Proprietary | |
| 24 | ERNIE 5.1 | Baidu | 1 489 | 1 475–1 503 | 1 905 | Proprietary |
| 25 | Hy3 | Tencent | 1 489 | 1 462–1 516 | 469 | Apache 2.0 |
| 26 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 489 | 1 470–1 508 | 941 | Proprietary |
| 27 | Gemini 3.5 Flash в режиме high | 1 489 | 1 475–1 502 | 2 186 | Proprietary | |
| 28 | GPT-5.5 · 5.5 | OpenAI | 1 488 | 1 478–1 499 | 3 706 | Proprietary |
| 29 | Claude Sonnet 5 в режиме high | Anthropic | 1 483 | 1 469–1 498 | 1 858 | Proprietary |
| 30 | Claude Sonnet 4.6 | Anthropic | 1 482 | 1 472–1 493 | 3 635 | Proprietary |
| 31 | GLM 5.2 в режиме max | Z.ai | 1 481 | 1 467–1 494 | 1 983 | MIT |
| 32 | Claude Opus 4.8 | Anthropic | 1 479 | 1 467–1 491 | 2 844 | Proprietary |
| 33 | Grok 4.5 | xAI | 1 479 | 1 463–1 495 | 1 496 | Proprietary |
| 34 | Kimi K2.6 | Moonshot AI | 1 479 | 1 465–1 492 | 2 046 | Modified MIT |
| 35 | Gemini 3.1 Pro Preview | 1 477 | 1 469–1 486 | 5 762 | Proprietary | |
| 36 | Qwen3.6 Max Preview | Alibaba Qwen | 1 476 | 1 445–1 507 | 342 | Proprietary |
| 37 | Gemini 3 Pro | 1 476 | 1 462–1 489 | 1 948 | Proprietary | |
| 38 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 476 | 1 460–1 491 | 1 482 | Proprietary |
| 39 | Qwen3.7 Plus | Alibaba Qwen | 1 474 | 1 460–1 487 | 2 055 | Proprietary |
| 40 | GLM 5.1 | Z.ai | 1 473 | 1 460–1 485 | 2 616 | MIT |
| 41 | Gemma 4 26B A4B | 1 471 | 1 439–1 503 | 296 | Apache 2.0 | |
| 42 | Kimi K2.5 · thinking | Moonshot AI | 1 471 | 1 461–1 481 | 3 708 | Modified MIT |
| 43 | Claude Opus 4.5 | Anthropic | 1 470 | 1 460–1 481 | 3 428 | Proprietary |
| 44 | Qwen3.7 Max | Alibaba Qwen | 1 468 | 1 429–1 508 | 228 | Proprietary |
| 45 | GPT-5.4 | OpenAI | 1 467 | 1 456–1 478 | 3 428 | Proprietary |
| 46 | Gemma 4 31B | 1 466 | 1 435–1 496 | 324 | Apache 2.0 | |
| 47 | Gemini 3 Flash Preview | 1 465 | 1 449–1 481 | 1 402 | Proprietary | |
| 48 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 465 | 1 449–1 481 | 1 510 | Proprietary |
| 49 | MiMo-V2 Pro | Xiaomi | 1 465 | 1 449–1 480 | 1 465 | Proprietary |
| 50 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 463 | 1 454–1 473 | 3 897 | Proprietary |
| 51 | Grok 4.6 в режиме high | xAI | 1 461 | 1 441–1 481 | 892 | Proprietary |
| 52 | Inkling | Thinking Machines Lab | 1 460 | 1 443–1 476 | 1 369 | Apache 2.0 |
| 53 | Nemotron 3 Ultra | NVIDIA | 1 459 | 1 436–1 482 | 654 | OpenMDW-1.1 |
| 54 | Qwen3.5 397B A17B | Alibaba Qwen | 1 455 | 1 446–1 465 | 4 319 | Apache 2.0 |
| 55 | Gemini 2.5 Pro | 1 455 | 1 448–1 463 | 6 591 | Proprietary | |
| 56 | DeepSeek V4 Pro 0423 | DeepSeek | 1 455 | 1 444–1 467 | 2 997 | MIT |
| 57 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 455 | 1 443–1 467 | 2 680 | MIT |
| 58 | Claude Opus 4.5 в режиме high | Anthropic | 1 454 | 1 439–1 469 | 1 533 | Proprietary |
| 59 | Qwen3.8 27B | Alibaba Qwen | 1 454 | 1 430–1 478 | 573 | Apache 2.0 |
| 60 | Muse Spark | Meta | 1 454 | 1 433–1 475 | 819 | Proprietary |
| 61 | GLM 5 | Z.ai | 1 451 | 1 435–1 466 | 1 474 | MIT |
| 62 | Qwen3.6 Plus | Alibaba Qwen | 1 451 | 1 439–1 463 | 2 626 | Proprietary |
| 63 | MiMo-V2.5 | Xiaomi | 1 450 | 1 437–1 462 | 2 434 | MIT |
| 64 | GLM 4.6 | Z.ai | 1 449 | 1 435–1 463 | 1 677 | MIT |
| 65 | GPT-5.1 в режиме high | OpenAI | 1 449 | 1 435–1 463 | 1 718 | Proprietary |
| 66 | MiniMax M3 | MiniMax | 1 447 | 1 435–1 460 | 2 616 | MiniMax Community License |
| 67 | MiMo-V2 Omni | Xiaomi | 1 446 | 1 427–1 465 | 970 | Proprietary |
| 68 | Qwen3 Max · preview | Alibaba Qwen | 1 446 | 1 430–1 462 | 1 306 | Proprietary |
| 69 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 446 | 1 418–1 474 | 446 | MIT |
| 70 | Claude Sonnet 4.5 | Anthropic | 1 446 | 1 436–1 456 | 3 991 | Proprietary |
| 71 | Grok 4.20 · beta-0309-reasoning | xAI | 1 446 | 1 435–1 456 | 3 342 | Proprietary |
| 72 | LongCat-Flash-Chat · chat | Meituan | 1 443 | 1 419–1 468 | 542 | MIT |
| 73 | GPT-5.2 в режиме high | OpenAI | 1 443 | 1 431–1 456 | 2 413 | Proprietary |
| 74 | Grok 4.20 Multi-Agent | xAI | 1 441 | 1 430–1 452 | 3 262 | Proprietary |
| 75 | Mistral Medium 3.5 | Mistral AI | 1 441 | 1 417–1 465 | 593 | Modified MIT |
| 76 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 441 | 1 432–1 449 | 4 937 | Apache 2.0 |
| 77 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 441 | 1 428–1 453 | 2 274 | MIT |
| 78 | GLM 5V Turbo | Z.ai | 1 441 | 1 415–1 466 | 518 | Proprietary |
| 79 | Hy3 preview | Tencent | 1 440 | 1 412–1 468 | 425 | tencent-hunyuan-community |
| 80 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 439 | 1 415–1 463 | 564 | MIT |
| 81 | Kimi K2.5 · instant | Moonshot AI | 1 438 | 1 407–1 469 | 330 | Modified MIT |
| 82 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 437 | 1 411–1 463 | 473 | Apache 2.0 |
| 83 | Kimi K2 Thinking | Moonshot AI | 1 437 | 1 426–1 448 | 2 962 | Modified MIT |
| 84 | Grok 4 Fast · chat | xAI | 1 436 | 1 404–1 468 | 310 | Proprietary |
| 85 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 436 | 1 421–1 451 | 1 613 | Proprietary |
| 86 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 436 | 1 407–1 465 | 372 | Apache 2.0 |
| 87 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 436 | 1 422–1 450 | 1 795 | MIT |
| 88 | Gemini 3.5 Flash Lite | 1 435 | 1 419–1 451 | 1 423 | Proprietary | |
| 89 | ERNIE 5.0 · 0110 | Baidu | 1 435 | 1 420–1 449 | 1 715 | Proprietary |
| 90 | Inkling Small | Thinking Machines Lab | 1 434 | 1 415–1 453 | 975 | Apache 2.0 |
| 91 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 434 | 1 422–1 446 | 2 379 | Proprietary |
| 92 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 434 | 1 416–1 452 | 1 034 | Apache 2.0 |
| 93 | Qwen3.5-27B | Alibaba Qwen | 1 434 | 1 418–1 449 | 1 535 | Apache 2.0 |
| 94 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 434 | 1 404–1 463 | 392 | MIT |
| 95 | Grok 4.20 · beta1 | xAI | 1 433 | 1 417–1 449 | 1 424 | Proprietary |
| 96 | MiniMax M2.7 | MiniMax | 1 433 | 1 423–1 444 | 3 552 | Modified MIT |
| 97 | Gemini 3 Flash Preview в режиме minimal | 1 433 | 1 424–1 443 | 4 413 | Proprietary | |
| 98 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 431 | 1 416–1 446 | 1 580 | Apache 2.0 |
| 99 | Seed 2.0 Pro | ByteDance Seed | 1 429 | 1 419–1 439 | 4 019 | Proprietary |
| 100 | DeepSeek V4 Flash 0423 | DeepSeek | 1 429 | 1 417–1 441 | 2 738 | MIT |
| 101 | GPT-5.2 Chat | OpenAI | 1 429 | 1 415–1 444 | 1 846 | Proprietary |
| 102 | ERNIE 5.0 · preview-1022 | Baidu | 1 429 | 1 395–1 463 | 265 | Proprietary |
| 103 | GPT-5.4 Mini в режиме high | OpenAI | 1 429 | 1 417–1 440 | 3 125 | Proprietary |
| 104 | Claude Haiku 4.5 | Anthropic | 1 429 | 1 421–1 437 | 6 596 | Proprietary |
| 105 | GLM 4.5 | Z.ai | 1 427 | 1 410–1 445 | 1 243 | MIT |
| 106 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 427 | 1 409–1 445 | 1 049 | Proprietary |
| 107 | Claude Opus 4.1 · 20250805 | Anthropic | 1 425 | 1 416–1 435 | 3 736 | Proprietary |
| 108 | Mistral Large 3 2512 | Mistral AI | 1 425 | 1 414–1 436 | 3 380 | Apache 2.0 |
| 109 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 424 | 1 412–1 436 | 2 595 | MIT |
| 110 | Grok 4 | xAI | 1 424 | 1 411–1 437 | 2 091 | Proprietary |
| 111 | GPT-5.1 | OpenAI | 1 422 | 1 409–1 435 | 2 038 | Proprietary |
| 112 | Gemini 2.5 Flash · flash | 1 422 | 1 415–1 429 | 6 769 | Proprietary | |
| 113 | o3 | OpenAI | 1 421 | 1 411–1 432 | 3 404 | Proprietary |
| 114 | Qwen3 32B | Alibaba Qwen | 1 419 | 1 388–1 451 | 282 | Apache 2.0 |
| 115 | GPT-5.2 | OpenAI | 1 419 | 1 409–1 429 | 4 120 | Proprietary |
| 116 | Step 3.5 Flash | StepFun | 1 419 | 1 407–1 430 | 2 903 | Apache 2.0 |
| 117 | MiniMax M2.1 | MiniMax | 1 419 | 1 397–1 440 | 651 | MIT |
| 118 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 419 | 1 404–1 433 | 1 628 | Proprietary | |
| 119 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 417 | 1 395–1 440 | 627 | MIT |
| 120 | Gemini 3.1 Flash Lite Preview | 1 417 | 1 406–1 427 | 3 309 | Proprietary | |
| 121 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 416 | 1 392–1 441 | 579 | Apache 2.0 |
| 122 | R1 0528 | DeepSeek | 1 416 | 1 397–1 435 | 915 | MIT |
| 123 | Mistral Medium 3.1 | Mistral AI | 1 415 | 1 407–1 424 | 4 790 | Proprietary |
| 124 | Grok 4.1 · 4.1-thinking | xAI | 1 415 | 1 404–1 426 | 3 010 | Proprietary |
| 125 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 415 | 1 389–1 441 | 466 | Proprietary |
| 126 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 414 | 1 402–1 427 | 2 301 | MIT |
| 127 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 414 | 1 394–1 434 | 831 | MIT |
| 128 | Qwen3.5-Flash | Alibaba Qwen | 1 414 | 1 402–1 425 | 3 057 | Proprietary |
| 129 | GPT-5.5 · 5.5-instant | OpenAI | 1 414 | 1 398–1 429 | 1 540 | Proprietary |
| 130 | ERNIE 5.0 · preview-1203 | Baidu | 1 414 | 1 387–1 440 | 486 | Proprietary |
| 131 | GLM 4.7 | Z.ai | 1 413 | 1 389–1 437 | 543 | MIT |
| 132 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 413 | 1 389–1 436 | 603 | Proprietary |
| 133 | GPT-4.5 Preview | OpenAI | 1 410 | 1 395–1 426 | 1 283 | Proprietary |
| 134 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 409 | 1 393–1 426 | 1 189 | Apache 2.0 |
| 135 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 407 | 1 392–1 422 | 1 592 | Apache 2.0 |
| 136 | Grok 4.1 · 4.1 | xAI | 1 407 | 1 396–1 418 | 3 364 | Proprietary |
| 137 | GPT-5.4 Nano в режиме high | OpenAI | 1 407 | 1 395–1 418 | 3 181 | Proprietary |
| 138 | Grok 3 | xAI | 1 406 | 1 394–1 418 | 2 424 | Proprietary |
| 139 | ChatGPT-4o (latest) | OpenAI | 1 405 | 1 397–1 414 | 4 728 | Proprietary |
| 140 | GPT-5 в режиме high | OpenAI | 1 404 | 1 389–1 418 | 1 735 | Proprietary |
| 141 | Nemotron 3 Super | NVIDIA | 1 403 | 1 376–1 431 | 409 | NVIDIA Open Model |
| 142 | Grok 4.1 Fast | xAI | 1 403 | 1 392–1 415 | 2 755 | Proprietary |
| 143 | Grok 4 Fast · reasoning | xAI | 1 403 | 1 384–1 422 | 928 | Proprietary |
| 144 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 402 | 1 390–1 414 | 2 267 | Apache 2.0 |
| 145 | Grok 4.3 | xAI | 1 400 | 1 389–1 411 | 3 578 | Proprietary |
| 146 | o3 Mini High | OpenAI | 1 398 | 1 385–1 412 | 1 711 | Proprietary |
| 147 | Kimi K2 0905 | Moonshot AI | 1 397 | 1 374–1 420 | 641 | Modified MIT |
| 148 | GPT-5 | OpenAI | 1 396 | 1 382–1 411 | 1 549 | Proprietary |
| 149 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 396 | 1 355–1 437 | 177 | Nvidia Open |
| 150 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 396 | 1 381–1 411 | 1 482 | Apache 2.0 |
| 151 | GLM 4.5 Air | Z.ai | 1 394 | 1 379–1 410 | 1 431 | MIT |
| 152 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 393 | 1 380–1 406 | 2 015 | Proprietary |
| 153 | Hunyuan T1 | Tencent | 1 392 | 1 353–1 431 | 219 | Proprietary |
| 154 | o1 · 2024-12-17 | OpenAI | 1 389 | 1 378–1 401 | 2 644 | Proprietary |
| 155 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 389 | 1 367–1 412 | 668 | Apache 2.0 |
| 156 | R1 | DeepSeek | 1 389 | 1 374–1 404 | 1 504 | MIT |
| 157 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 386 | 1 359–1 413 | 418 | MIT |
| 158 | gpt-oss-120b | OpenAI | 1 384 | 1 369–1 399 | 1 504 | Apache 2.0 |
| 159 | GPT-5.3 Chat | OpenAI | 1 381 | 1 367–1 396 | 1 856 | Proprietary |
| 160 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 380 | 1 367–1 394 | 1 868 | Proprietary |
| 161 | GPT-5 Mini в режиме high | OpenAI | 1 380 | 1 364–1 397 | 1 266 | Proprietary |
| 162 | o1 · preview | OpenAI | 1 380 | 1 370–1 391 | 3 876 | Proprietary |
| 163 | MiniMax M2.5 | MiniMax | 1 380 | 1 367–1 393 | 2 239 | Modified MIT |
| 164 | INTELLECT-3 | Prime Intellect | 1 379 | 1 343–1 415 | 238 | MIT |
| 165 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 379 | 1 366–1 392 | 2 085 | Proprietary | |
| 166 | o3 Mini | OpenAI | 1 379 | 1 370–1 388 | 4 351 | Proprietary |
| 167 | Claude Opus 4 · 20250514 | Anthropic | 1 378 | 1 366–1 389 | 2 586 | Proprietary |
| 168 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 378 | 1 355–1 400 | 668 | NVIDIA Open Model |
| 169 | o4 Mini | OpenAI | 1 377 | 1 366–1 388 | 2 747 | Proprietary |
| 170 | Muse Glimmer 30B | Meta | 1 377 | 1 337–1 416 | 229 | Apache-2.0 |
| 171 | Step 3 | StepFun | 1 376 | 1 342–1 409 | 299 | Apache 2.0 |
| 172 | GLM 4.7 Flash | Z.ai | 1 375 | 1 349–1 401 | 476 | MIT |
| 173 | Nova 2 Lite | Amazon | 1 373 | 1 349–1 397 | 589 | Proprietary |
| 174 | MiniMax M2 | MiniMax | 1 373 | 1 339–1 407 | 289 | Apache 2.0 |
| 175 | Ling-flash-2.0 | inclusionAI | 1 369 | 1 339–1 398 | 359 | MIT |
| 176 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 368 | 1 355–1 381 | 1 957 | Proprietary | |
| 177 | Grok 3 Mini в режиме high | xAI | 1 368 | 1 349–1 386 | 945 | Proprietary |
| 178 | Qwen2.5 Max | Alibaba Qwen | 1 367 | 1 357–1 378 | 2 940 | Proprietary |
| 179 | o1-mini | OpenAI | 1 366 | 1 358–1 375 | 6 318 | Proprietary |
| 180 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 365 | 1 349–1 381 | 1 439 | Apache 2.0 |
| 181 | DeepSeek V3 0324 | DeepSeek | 1 365 | 1 354–1 375 | 2 913 | MIT |
| 182 | Qwen3 30B A3B | Alibaba Qwen | 1 363 | 1 349–1 378 | 1 565 | Apache 2.0 |
| 183 | MiniMax M1 | MiniMax | 1 361 | 1 348–1 375 | 1 732 | Apache 2.0 |
| 184 | Hunyuan TurboS · 20250416 | Tencent | 1 360 | 1 340–1 380 | 795 | Proprietary |
| 185 | Grok 3 Mini | xAI | 1 359 | 1 344–1 374 | 1 401 | Proprietary |
| 186 | GPT-4.1 | OpenAI | 1 359 | 1 348–1 369 | 3 034 | Proprietary |
| 187 | Claude Sonnet 4 · 20250514 | Anthropic | 1 359 | 1 346–1 371 | 2 273 | Proprietary |
| 188 | Nemotron 3.5 Lightning | NVIDIA | 1 358 | 1 327–1 388 | 424 | OpenMDW-1.1 |
| 189 | Ring-flash-2.0 | inclusionAI | 1 357 | 1 326–1 388 | 351 | MIT |
| 190 | Gemini 2.0 Flash | 1 356 | 1 347–1 366 | 3 733 | Proprietary | |
| 191 | QwQ 32B · 32b | Alibaba Qwen | 1 356 | 1 342–1 370 | 1 561 | Apache 2.0 |
| 192 | Kimi K2 0711 | Moonshot AI | 1 356 | 1 341–1 371 | 1 553 | Modified MIT |
| 193 | Trinity Large Thinking | Arcee AI | 1 355 | 1 339–1 370 | 1 641 | Apache 2.0 |
| 194 | Trinity Large | Arcee AI | 1 355 | 1 339–1 370 | 1 631 | Apache 2.0 |
| 195 | Mistral Medium 3 | Mistral AI | 1 353 | 1 341–1 366 | 2 167 | Proprietary |
| 196 | OLMo 3.1 32B Instruct | Ai2 | 1 352 | 1 323–1 380 | 411 | Apache 2.0 |
| 197 | Mistral Small 3.2 24B | Mistral AI | 1 351 | 1 333–1 370 | 981 | Apache 2.0 |
| 198 | GPT-4.1 Mini | OpenAI | 1 346 | 1 334–1 358 | 2 455 | Proprietary |
| 199 | Step-1o Turbo | StepFun | 1 344 | 1 320–1 369 | 566 | Proprietary |
| 200 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 344 | 1 307–1 381 | 207 | Nvidia |
| 201 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 338 | 1 327–1 350 | 2 537 | Proprietary |
| 202 | Gemma 3 12B | 1 336 | 1 306–1 366 | 333 | Gemma | |
| 203 | GLM 4.5V | Z.ai | 1 334 | 1 297–1 371 | 228 | MIT |
| 204 | Qwen-Plus | Alibaba Qwen | 1 332 | 1 311–1 354 | 627 | Proprietary |
| 205 | OLMo 3 32B Think | Ai2 | 1 328 | 1 288–1 367 | 198 | Apache 2.0 |
| 206 | Gemma 3 27B | 1 327 | 1 316–1 337 | 3 156 | Gemma | |
| 207 | Gemini 1.5 Pro · 1.5-pro-002 | 1 324 | 1 317–1 332 | 6 506 | Proprietary | |
| 208 | GPT-5 Nano в режиме high | OpenAI | 1 324 | 1 296–1 353 | 435 | Proprietary |
| 209 | Granite 4.1 8B | IBM Granite | 1 320 | 1 279–1 361 | 214 | Apache 2.0 |
| 210 | Gemini 2.0 Flash-Lite | 1 318 | 1 306–1 329 | 2 399 | Proprietary | |
| 211 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 314 | 1 304–1 325 | 3 154 | Proprietary |
| 212 | DeepSeek V3 | DeepSeek | 1 314 | 1 302–1 326 | 2 326 | DeepSeek |
| 213 | Command A | Cohere | 1 310 | 1 301–1 320 | 3 607 | CC-BY-NC-4.0 |
| 214 | Step-2 16k | StepFun | 1 310 | 1 288–1 332 | 543 | Proprietary |
| 215 | OLMo 3.1 32B Think | Ai2 | 1 307 | 1 277–1 338 | 342 | Apache 2.0 |
| 216 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 307 | 1 300–1 314 | 8 593 | Proprietary |
| 217 | Yi-Lightning | 01.AI | 1 306 | 1 296–1 317 | 3 206 | Proprietary |
| 218 | Athene V2 Chat | Nexusflow | 1 305 | 1 295–1 315 | 2 942 | NexusFlow |
| 219 | Qwen2.5 Plus | Alibaba Qwen | 1 305 | 1 290–1 320 | 1 265 | Proprietary |
| 220 | gpt-oss-20b | OpenAI | 1 304 | 1 280–1 329 | 574 | Apache 2.0 |
| 221 | Hunyuan Large | Tencent | 1 304 | 1 278–1 330 | 431 | Proprietary |
| 222 | Hunyuan Large Vision | Tencent | 1 302 | 1 273–1 331 | 361 | Proprietary |
| 223 | Hunyuan Turbo | Tencent | 1 298 | 1 265–1 332 | 216 | Proprietary |
| 224 | Llama 4 Maverick | Meta | 1 294 | 1 283–1 306 | 2 557 | Llama 4 |
| 225 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 294 | 1 286–1 301 | 10 038 | Proprietary |
| 226 | GLM-4-Plus · plus-0111 | Z.ai | 1 291 | 1 270–1 312 | 638 | Proprietary |
| 227 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 290 | 1 282–1 299 | 4 565 | Qwen |
| 228 | Hunyuan TurboS · 20250226 | Tencent | 1 290 | 1 258–1 322 | 224 | Proprietary |
| 229 | Hunyuan Standard · 2025-02-10 | Tencent | 1 288 | 1 263–1 314 | 441 | Proprietary |
| 230 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 287 | 1 269–1 306 | 874 | Llama 3.1 |
| 231 | GPT-4o (2024-05-13) | OpenAI | 1 284 | 1 277–1 292 | 13 306 | Proprietary |
| 232 | Llama 4 Scout | Meta | 1 284 | 1 270–1 298 | 1 797 | Llama |
| 233 | Grok 2 | xAI | 1 284 | 1 276–1 291 | 7 604 | Proprietary |
| 234 | GPT-4o (2024-08-06) | OpenAI | 1 283 | 1 275–1 292 | 5 805 | Proprietary |
| 235 | GLM-4-Plus · plus | Z.ai | 1 283 | 1 272–1 293 | 2 991 | Proprietary |
| 236 | Qwen Max | Alibaba Qwen | 1 279 | 1 266–1 292 | 1 804 | Qwen |
| 237 | Gemini 1.5 Flash · 002 | 1 279 | 1 270–1 288 | 4 114 | Proprietary | |
| 238 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 278 | 1 261–1 296 | 893 | DeepSeek |
| 239 | Gemini 1.5 Pro · 1.5-pro-001 | 1 276 | 1 268–1 284 | 9 328 | Proprietary | |
| 240 | Llama 3.1 405B Instruct · fp8 | Meta | 1 276 | 1 268–1 284 | 7 242 | Llama 3.1 Community |
| 241 | Gemini 1.5 Pro · advanced-0514 | 1 276 | 1 266–1 286 | 5 787 | Proprietary | |
| 242 | Llama 3.1 405B Instruct · bf16 | Meta | 1 276 | 1 267–1 284 | 4 366 | Llama 3.1 Community |
| 243 | GPT-4.1 Nano | OpenAI | 1 274 | 1 250–1 298 | 534 | Proprietary |
| 244 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 273 | 1 263–1 284 | 3 039 | DeepSeek |
| 245 | Gemma 3n E4B | 1 273 | 1 258–1 289 | 1 393 | Gemma | |
| 246 | Granite 4.0 H Small | IBM Granite | 1 273 | 1 236–1 309 | 266 | Apache 2.0 |
| 247 | Grok 2 Mini | xAI | 1 273 | 1 265–1 281 | 6 253 | Proprietary |
| 248 | GPT-4o-mini (2024-07-18) | OpenAI | 1 271 | 1 264–1 279 | 7 963 | Proprietary |
| 249 | Mistral Small 3.1 24B | Mistral AI | 1 270 | 1 257–1 284 | 1 908 | Apache 2.0 |
| 250 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 270 | 1 262–1 278 | 11 549 | Proprietary |
| 251 | GPT-4 Turbo | OpenAI | 1 268 | 1 260–1 276 | 11 796 | Proprietary |
| 252 | Llama 3.3 70B Instruct | Meta | 1 267 | 1 259–1 275 | 5 226 | Llama-3.3 |
| 253 | Mistral Large 2407 | Mistral AI | 1 265 | 1 257–1 274 | 5 745 | Mistral Research |
| 254 | Claude 3 Opus | Anthropic | 1 264 | 1 257–1 270 | 22 970 | Proprietary |
| 255 | Nova Pro 1.0 | Amazon | 1 261 | 1 251–1 272 | 2 650 | Proprietary |
| 256 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 261 | 1 253–1 269 | 10 830 | Proprietary |
| 257 | Mistral Large · 2411 | Mistral AI | 1 257 | 1 247–1 267 | 3 037 | MRL |
| 258 | Gemma 3 4B | 1 256 | 1 226–1 286 | 365 | Gemma | |
| 259 | Phi 4 | Microsoft | 1 254 | 1 242–1 265 | 2 350 | MIT |
| 260 | Llama 3.1 70B Instruct | Meta | 1 253 | 1 245–1 262 | 6 524 | Llama 3.1 Community |
| 261 | Magistral Medium | Mistral AI | 1 252 | 1 226–1 279 | 570 | Proprietary |
| 262 | Claude 3.5 Haiku | Anthropic | 1 249 | 1 241–1 257 | 5 743 | Proprietary |
| 263 | Llama 3.1 Tulu 3 70B | Ai2 | 1 246 | 1 218–1 274 | 349 | Llama 3.1 |
| 264 | Hunyuan Standard · 256k | Tencent | 1 243 | 1 212–1 274 | 292 | Proprietary |
| 265 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 243 | 1 223–1 264 | 619 | Apache 2.0 |
| 266 | Mistral Small 3 | Mistral AI | 1 243 | 1 229–1 257 | 1 447 | Apache 2.0 |
| 267 | Reka Core | Reka AI | 1 239 | 1 223–1 255 | 1 039 | Proprietary |
| 268 | Gemini 1.5 Flash · 001 | 1 238 | 1 230–1 246 | 7 315 | Proprietary | |
| 269 | Nova Lite 1.0 | Amazon | 1 234 | 1 223–1 246 | 2 173 | Proprietary |
| 270 | Athene 70B | Nexusflow | 1 233 | 1 222–1 244 | 2 559 | CC-BY-NC-4.0 |
| 271 | DeepSeek Coder V2 | DeepSeek | 1 233 | 1 219–1 247 | 1 669 | DeepSeek License |
| 272 | Qwen2 72B Instruct | Alibaba Qwen | 1 231 | 1 221–1 241 | 4 235 | Qianwen LICENSE |
| 273 | GLM-4 | Z.ai | 1 230 | 1 213–1 246 | 1 067 | Proprietary |
| 274 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 228 | 1 203–1 254 | 422 | Llama 3.1 |
| 275 | GPT-4 · 0314 | OpenAI | 1 228 | 1 218–1 238 | 6 112 | Proprietary |
| 276 | Jamba 1.5 Large | AI21 Labs | 1 228 | 1 211–1 244 | 962 | Jamba Open |
| 277 | Gemini 1.5 Flash-8B | 1 226 | 1 217–1 235 | 4 171 | Proprietary | |
| 278 | QwQ 32B · 32b-preview | Alibaba Qwen | 1 221 | 1 195–1 248 | 422 | Apache 2.0 |
| 279 | Gemma 2 27B | 1 221 | 1 214–1 229 | 8 869 | Gemma license | |
| 280 | Nemotron-4 340B Instruct | NVIDIA | 1 216 | 1 203–1 228 | 2 133 | NVIDIA Open Model |
| 281 | Llama 3 70B Instruct | Meta | 1 214 | 1 206–1 221 | 18 728 | Llama 3 Community |
| 282 | Claude 3 Sonnet | Anthropic | 1 213 | 1 205–1 221 | 12 413 | Proprietary |
| 283 | Nova Micro 1.0 | Amazon | 1 212 | 1 200–1 224 | 2 016 | Proprietary |
| 284 | Aya Expanse 32B | Cohere | 1 208 | 1 198–1 218 | 3 239 | CC-BY-NC-4.0 |
| 285 | Reka Flash (2024-09) | Reka AI | 1 202 | 1 187–1 217 | 1 109 | Proprietary |
| 286 | GPT-4 · 0613 | OpenAI | 1 201 | 1 192–1 210 | 9 862 | Proprietary |
| 287 | Command R+ (08-2024) | Cohere | 1 198 | 1 183–1 213 | 1 243 | CC-BY-NC-4.0 |
| 288 | Gemma 2 9B IT SimPO | Princeton NLP | 1 196 | 1 180–1 212 | 1 139 | MIT |
| 289 | OLMo 2 32B Instruct | Ai2 | 1 195 | 1 165–1 225 | 336 | Apache-2.0 |
| 290 | Gemma 2 9B | 1 193 | 1 185–1 201 | 6 297 | Gemma license | |
| 291 | Llama 3.1 Tulu 3 8B | Ai2 | 1 193 | 1 163–1 222 | 304 | Llama 3.1 |
| 292 | Qwen1.5 110B Chat | Alibaba Qwen | 1 190 | 1 178–1 202 | 2 860 | Qianwen LICENSE |
| 293 | Mistral Large · 2402 | Mistral AI | 1 190 | 1 180–1 199 | 7 028 | Proprietary |
| 294 | Claude 3 Haiku | Anthropic | 1 187 | 1 180–1 195 | 13 491 | Proprietary |
| 295 | Granite 3.1 2B Instruct | IBM Granite | 1 186 | 1 158–1 215 | 341 | Apache 2.0 |
| 296 | Yi-1.5 34B Chat | 01.AI | 1 185 | 1 174–1 197 | 2 568 | Apache-2.0 |
| 297 | Ministral 8B (2410) | Mistral AI | 1 184 | 1 162–1 206 | 543 | MRL |
| 298 | InternLM2.5 20B Chat | InternLM | 1 181 | 1 165–1 197 | 1 086 | Other |
| 299 | Granite 3.1 8B Instruct | IBM Granite | 1 180 | 1 150–1 211 | 320 | Apache 2.0 |
| 300 | Llama 3.1 8B Instruct | Meta | 1 180 | 1 172–1 189 | 5 985 | Llama 3.1 Community |
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.