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 «следование инструкциям», 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 | 1 523 | 1 518–1 529 | 22 942 | Proprietary |
| 2 | Claude Opus 5 в режиме high | Anthropic | 1 517 | 1 511–1 524 | 15 145 | Proprietary |
| 3 | Claude Opus 5 в режиме max | Anthropic | 1 516 | 1 508–1 524 | 7 148 | Proprietary |
| 4 | Claude Fable 5.1 в режиме max | Anthropic | 1 513 | 1 499–1 528 | 1 773 | Proprietary |
| 5 | Claude Opus 4.6 | Anthropic | 1 511 | 1 506–1 516 | 25 156 | Proprietary |
| 6 | Claude Fable 5 | Anthropic | 1 505 | 1 498–1 511 | 10 641 | Proprietary |
| 7 | Claude Opus 4.7 в режиме high | Anthropic | 1 498 | 1 492–1 503 | 20 821 | Proprietary |
| 8 | Gemini 3.8 Flash в режиме high | 1 493 | 1 479–1 507 | 1 804 | Proprietary | |
| 9 | Claude Opus 4.7 | Anthropic | 1 488 | 1 482–1 494 | 21 342 | Proprietary |
| 10 | Gemini 3.7 Flash в режиме high | 1 484 | 1 471–1 498 | 2 035 | Proprietary | |
| 11 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 482 | 1 474–1 491 | 5 969 | Proprietary |
| 12 | GLM 5.3 в режиме max | Z.ai | 1 481 | 1 471–1 491 | 3 825 | MIT |
| 13 | Muse Spark 1.3 в режиме max | Meta | 1 479 | 1 465–1 494 | 1 693 | Proprietary |
| 14 | Kimi K3 в режиме max | Moonshot AI | 1 478 | 1 471–1 486 | 6 985 | Kimi K3 license |
| 15 | GPT-5.5 в режиме high | OpenAI | 1 478 | 1 473–1 484 | 22 336 | Proprietary |
| 16 | MiMo-V2.5-Pro | Xiaomi | 1 478 | 1 472–1 483 | 20 161 | MIT |
| 17 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 477 | 1 470–1 484 | 9 363 | Proprietary |
| 18 | Claude Opus 4.5 в режиме high | Anthropic | 1 477 | 1 470–1 484 | 9 457 | Proprietary |
| 19 | Claude Sonnet 4.6 | Anthropic | 1 477 | 1 472–1 482 | 22 010 | Proprietary |
| 20 | GLM 5.3 Flash | Z.ai | 1 476 | 1 466–1 487 | 3 554 | MIT |
| 21 | Claude Opus 4.8 в режиме high | Anthropic | 1 476 | 1 470–1 482 | 18 394 | Proprietary |
| 22 | GPT-5.5 · 5.5 | OpenAI | 1 473 | 1 467–1 478 | 23 028 | Proprietary |
| 23 | Claude Opus 4.5 | Anthropic | 1 473 | 1 467–1 478 | 20 732 | Proprietary |
| 24 | Muse Spark 1.2 в режиме xhigh | Meta | 1 471 | 1 454–1 488 | 1 174 | Proprietary |
| 25 | GPT-5.4 в режиме high | OpenAI | 1 470 | 1 465–1 476 | 20 257 | Proprietary |
| 26 | Qwen3.5 Max | Alibaba Qwen | 1 469 | 1 461–1 476 | 6 955 | Proprietary |
| 27 | Gemini 3.6 Flash в режиме high | 1 467 | 1 460–1 474 | 9 558 | Proprietary | |
| 28 | Gemini 3.1 Pro Preview | 1 467 | 1 462–1 471 | 36 184 | Proprietary | |
| 29 | Gemini 3.5 Flash в режиме high | 1 466 | 1 460–1 472 | 13 634 | Proprietary | |
| 30 | Claude Opus 4.8 | Anthropic | 1 462 | 1 456–1 467 | 18 879 | Proprietary |
| 31 | Gemini 3.5 Flash в режиме medium | 1 461 | 1 455–1 467 | 13 218 | Proprietary | |
| 32 | ERNIE 5.1 | Baidu | 1 460 | 1 454–1 467 | 11 716 | Proprietary |
| 33 | GLM 5.2 в режиме max | Z.ai | 1 460 | 1 453–1 466 | 12 822 | MIT |
| 34 | Claude Sonnet 4.5 | Anthropic | 1 459 | 1 454–1 464 | 23 296 | Proprietary |
| 35 | Qwen3.7 Max | Alibaba Qwen | 1 458 | 1 441–1 475 | 1 283 | Proprietary |
| 36 | Gemini 3 Pro | 1 458 | 1 452–1 465 | 11 193 | Proprietary | |
| 37 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 458 | 1 453–1 462 | 23 592 | Proprietary |
| 38 | Muse Spark 1.1 | Meta | 1 456 | 1 449–1 463 | 9 923 | Proprietary |
| 39 | GPT-5.4 | OpenAI | 1 455 | 1 450–1 461 | 21 724 | Proprietary |
| 40 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 453 | 1 446–1 460 | 9 959 | Proprietary |
| 41 | Claude Sonnet 5 в режиме high | Anthropic | 1 450 | 1 444–1 457 | 12 431 | Proprietary |
| 42 | GLM 5.1 | Z.ai | 1 450 | 1 445–1 456 | 17 018 | MIT |
| 43 | Grok 4.5 | xAI | 1 449 | 1 443–1 456 | 10 848 | Proprietary |
| 44 | Kimi K2.6 | Moonshot AI | 1 449 | 1 442–1 455 | 12 215 | Modified MIT |
| 45 | DeepSeek V4 Pro 0423 | DeepSeek | 1 445 | 1 439–1 450 | 18 649 | MIT |
| 46 | MiMo-V2 Pro | Xiaomi | 1 444 | 1 436–1 451 | 8 027 | Proprietary |
| 47 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 443 | 1 432–1 453 | 3 162 | MIT |
| 48 | Muse Spark | Meta | 1 442 | 1 432–1 451 | 4 494 | Proprietary |
| 49 | GPT-5.1 в режиме high | OpenAI | 1 441 | 1 435–1 448 | 10 718 | Proprietary |
| 50 | GPT-6 Astra в режиме max | OpenAI | 1 439 | 1 419–1 460 | 846 | Proprietary |
| 51 | Qwen3.7 Plus | Alibaba Qwen | 1 439 | 1 433–1 445 | 13 902 | Proprietary |
| 52 | Gemini 3 Flash Preview | 1 438 | 1 431–1 445 | 8 117 | Proprietary | |
| 53 | Gemini 2.5 Pro | 1 437 | 1 433–1 441 | 34 174 | Proprietary | |
| 54 | Qwen3.6 Max Preview | Alibaba Qwen | 1 437 | 1 423–1 451 | 1 782 | Proprietary |
| 55 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 437 | 1 431–1 442 | 12 909 | Proprietary |
| 56 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 436 | 1 430–1 442 | 17 705 | MIT |
| 57 | Qwen3.8 27B | Alibaba Qwen | 1 436 | 1 426–1 446 | 3 778 | Apache 2.0 |
| 58 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 435 | 1 428–1 442 | 10 129 | Proprietary |
| 59 | MiMo-V2.5 | Xiaomi | 1 434 | 1 428–1 440 | 14 956 | MIT |
| 60 | Claude Opus 4.1 · 20250805 | Anthropic | 1 434 | 1 429–1 438 | 20 331 | Proprietary |
| 61 | Grok 4.6 в режиме high | xAI | 1 433 | 1 425–1 442 | 5 549 | Proprietary |
| 62 | MiniMax M3 | MiniMax | 1 432 | 1 427–1 438 | 16 812 | MiniMax Community License |
| 63 | Kimi K2.5 · instant | Moonshot AI | 1 432 | 1 420–1 444 | 2 216 | Modified MIT |
| 64 | Gemma 4 31B | 1 432 | 1 418–1 446 | 1 707 | Apache 2.0 | |
| 65 | Kimi K2.5 · thinking | Moonshot AI | 1 431 | 1 426–1 436 | 22 600 | Modified MIT |
| 66 | MiMo-V2 Omni | Xiaomi | 1 428 | 1 419–1 436 | 6 577 | Proprietary |
| 67 | GLM 5 | Z.ai | 1 428 | 1 421–1 434 | 8 967 | MIT |
| 68 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 426 | 1 407–1 444 | 933 | Proprietary |
| 69 | Qwen3.6 Plus | Alibaba Qwen | 1 425 | 1 419–1 431 | 15 410 | Proprietary |
| 70 | Qwen3.5 397B A17B | Alibaba Qwen | 1 425 | 1 420–1 429 | 25 939 | Apache 2.0 |
| 71 | Hy3 | Tencent | 1 423 | 1 412–1 435 | 2 839 | Apache 2.0 |
| 72 | GLM 5V Turbo | Z.ai | 1 421 | 1 411–1 432 | 3 244 | Proprietary |
| 73 | DeepSeek V4 Flash 0423 | DeepSeek | 1 421 | 1 415–1 427 | 16 752 | MIT |
| 74 | Qwen3 Max · preview | Alibaba Qwen | 1 421 | 1 413–1 428 | 7 243 | Proprietary |
| 75 | Gemma 4 26B A4B | 1 421 | 1 407–1 435 | 1 628 | Apache 2.0 | |
| 76 | Inkling | Thinking Machines Lab | 1 419 | 1 412–1 426 | 9 291 | Apache 2.0 |
| 77 | Grok 4.20 Multi-Agent | xAI | 1 419 | 1 414–1 425 | 20 355 | Proprietary |
| 78 | Grok 4.20 · beta-0309-reasoning | xAI | 1 419 | 1 413–1 424 | 20 957 | Proprietary |
| 79 | Gemini 3.5 Flash Lite | 1 419 | 1 412–1 426 | 9 372 | Proprietary | |
| 80 | GPT-5.1 | OpenAI | 1 418 | 1 412–1 424 | 11 857 | Proprietary |
| 81 | GPT-5.2 Chat | OpenAI | 1 417 | 1 411–1 423 | 11 054 | Proprietary |
| 82 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 417 | 1 411–1 423 | 16 773 | MIT |
| 83 | Nemotron 3 Ultra | NVIDIA | 1 416 | 1 406–1 427 | 3 605 | OpenMDW-1.1 |
| 84 | Claude Haiku 4.5 | Anthropic | 1 415 | 1 411–1 419 | 40 629 | Proprietary |
| 85 | Grok 4.20 · beta1 | xAI | 1 415 | 1 408–1 422 | 8 415 | Proprietary |
| 86 | Gemini 3 Flash Preview в режиме minimal | 1 414 | 1 409–1 419 | 27 812 | Proprietary | |
| 87 | Mistral Medium 3.5 | Mistral AI | 1 414 | 1 404–1 424 | 3 549 | Modified MIT |
| 88 | ERNIE 5.0 · 0110 | Baidu | 1 414 | 1 408–1 420 | 10 515 | Proprietary |
| 89 | GLM 4.7 | Z.ai | 1 413 | 1 402–1 423 | 3 207 | MIT |
| 90 | Seed 2.0 Pro | ByteDance Seed | 1 413 | 1 408–1 418 | 24 409 | Proprietary |
| 91 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 413 | 1 407–1 418 | 12 970 | MIT |
| 92 | GLM 4.6 | Z.ai | 1 411 | 1 404–1 417 | 10 003 | MIT |
| 93 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 410 | 1 403–1 417 | 9 142 | Proprietary |
| 94 | GPT-5.2 в режиме high | OpenAI | 1 409 | 1 403–1 414 | 14 524 | Proprietary |
| 95 | Grok 3 | xAI | 1 409 | 1 402–1 415 | 9 628 | Proprietary |
| 96 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 409 | 1 405–1 413 | 26 989 | Apache 2.0 |
| 97 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 407 | 1 401–1 413 | 11 200 | MIT |
| 98 | GPT-5.5 · 5.5-instant | OpenAI | 1 407 | 1 400–1 414 | 9 141 | Proprietary |
| 99 | DeepSeek V3.1 Terminus · terminus-thinking | DeepSeek | 1 407 | 1 387–1 427 | 852 | MIT |
| 100 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 407 | 1 396–1 418 | 2 914 | Apache 2.0 |
| 101 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 406 | 1 399–1 413 | 9 084 | Proprietary |
| 102 | Gemini 2.5 Flash · flash | 1 405 | 1 402–1 409 | 33 886 | Proprietary | |
| 103 | GPT-5.4 Mini в режиме high | OpenAI | 1 405 | 1 399–1 410 | 20 176 | Proprietary |
| 104 | GLM 4.5 | Z.ai | 1 404 | 1 396–1 412 | 6 116 | MIT |
| 105 | GPT-4.5 Preview | OpenAI | 1 404 | 1 396–1 412 | 5 501 | Proprietary |
| 106 | ChatGPT-4o (latest) | OpenAI | 1 404 | 1 399–1 408 | 22 553 | Proprietary |
| 107 | MiniMax M2.7 | MiniMax | 1 404 | 1 398–1 409 | 22 853 | Modified MIT |
| 108 | ERNIE 5.0 · preview-1203 | Baidu | 1 403 | 1 392–1 415 | 2 615 | Proprietary |
| 109 | GPT-5.2 | OpenAI | 1 403 | 1 398–1 408 | 25 926 | Proprietary |
| 110 | Kimi K2 Thinking | Moonshot AI | 1 403 | 1 398–1 408 | 17 907 | Modified MIT |
| 111 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 401 | 1 391–1 412 | 3 319 | MIT |
| 112 | Mistral Large 3 2512 | Mistral AI | 1 401 | 1 396–1 405 | 21 641 | Apache 2.0 |
| 113 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 400 | 1 394–1 407 | 9 070 | Apache 2.0 |
| 114 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 400 | 1 389–1 412 | 2 832 | MIT |
| 115 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 400 | 1 389–1 412 | 2 528 | MIT |
| 116 | Grok 4.1 · 4.1 | xAI | 1 400 | 1 395–1 405 | 18 647 | Proprietary |
| 117 | Hunyuan Vision 1.5 | Tencent | 1 399 | 1 377–1 421 | 626 | Proprietary |
| 118 | LongCat-Flash-Chat · chat | Meituan | 1 398 | 1 387–1 409 | 2 931 | MIT |
| 119 | Mistral Medium 3.1 | Mistral AI | 1 398 | 1 394–1 402 | 26 711 | Proprietary |
| 120 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 398 | 1 387–1 409 | 2 620 | Proprietary |
| 121 | MiniMax M2.1 | MiniMax | 1 398 | 1 389–1 406 | 4 439 | MIT |
| 122 | Inkling Small | Thinking Machines Lab | 1 397 | 1 389–1 404 | 6 904 | Apache 2.0 |
| 123 | Grok 4.1 · 4.1-thinking | xAI | 1 395 | 1 390–1 401 | 18 420 | Proprietary |
| 124 | ERNIE 5.0 · preview-1022 | Baidu | 1 395 | 1 380–1 411 | 1 318 | Proprietary |
| 125 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 394 | 1 388–1 401 | 9 128 | Proprietary | |
| 126 | Qwen3.5-27B | Alibaba Qwen | 1 392 | 1 385–1 399 | 8 654 | Apache 2.0 |
| 127 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 392 | 1 387–1 398 | 14 027 | MIT |
| 128 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 391 | 1 372–1 410 | 925 | Proprietary |
| 129 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 389 | 1 379–1 399 | 3 690 | MIT |
| 130 | GPT-5 в режиме high | OpenAI | 1 389 | 1 381–1 396 | 8 259 | Proprietary |
| 131 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 388 | 1 381–1 396 | 6 321 | Apache 2.0 |
| 132 | Hy3 preview | Tencent | 1 388 | 1 375–1 400 | 2 228 | tencent-hunyuan-community |
| 133 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 387 | 1 380–1 394 | 6 736 | Proprietary |
| 134 | Grok 4 | xAI | 1 387 | 1 381–1 393 | 10 675 | Proprietary |
| 135 | Grok 4 Fast · chat | xAI | 1 387 | 1 372–1 401 | 1 673 | Proprietary |
| 136 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 385 | 1 373–1 398 | 2 086 | Apache 2.0 |
| 137 | Step 3.5 Flash | StepFun | 1 385 | 1 380–1 391 | 18 175 | Apache 2.0 |
| 138 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 383 | 1 365–1 401 | 925 | Proprietary |
| 139 | Grok 4 Fast · reasoning | xAI | 1 383 | 1 375–1 391 | 5 210 | Proprietary |
| 140 | GPT-5 | OpenAI | 1 382 | 1 376–1 389 | 8 127 | Proprietary |
| 141 | R1 0528 | DeepSeek | 1 381 | 1 372–1 391 | 3 997 | MIT |
| 142 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 380 | 1 374–1 387 | 9 370 | Apache 2.0 |
| 143 | DeepSeek V3.1 Terminus · terminus | DeepSeek | 1 380 | 1 362–1 398 | 1 020 | MIT |
| 144 | Muse Glimmer 30B | Meta | 1 378 | 1 361–1 394 | 1 266 | Apache-2.0 |
| 145 | GPT-5.3 Chat | OpenAI | 1 377 | 1 371–1 384 | 10 615 | Proprietary |
| 146 | Grok 4.1 Fast | xAI | 1 376 | 1 370–1 381 | 16 533 | Proprietary |
| 147 | Hunyuan T1 | Tencent | 1 376 | 1 358–1 393 | 1 099 | Proprietary |
| 148 | Gemini 3.1 Flash Lite Preview | 1 376 | 1 370–1 381 | 20 001 | Proprietary | |
| 149 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 375 | 1 368–1 382 | 8 459 | Proprietary |
| 150 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 375 | 1 362–1 387 | 2 127 | Apache 2.0 |
| 151 | Qwen3.5-Flash | Alibaba Qwen | 1 375 | 1 369–1 380 | 18 914 | Proprietary |
| 152 | Claude Opus 4 · 20250514 | Anthropic | 1 374 | 1 367–1 380 | 10 616 | Proprietary |
| 153 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 373 | 1 362–1 384 | 2 966 | MIT |
| 154 | o3 | OpenAI | 1 368 | 1 362–1 373 | 15 386 | Proprietary |
| 155 | o1 · 2024-12-17 | OpenAI | 1 368 | 1 361–1 374 | 10 246 | Proprietary |
| 156 | GPT-4.1 | OpenAI | 1 367 | 1 361–1 373 | 13 178 | Proprietary |
| 157 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 366 | 1 359–1 373 | 8 056 | Proprietary | |
| 158 | Grok 4.3 | xAI | 1 366 | 1 360–1 371 | 23 367 | Proprietary |
| 159 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 364 | 1 354–1 375 | 3 040 | Proprietary |
| 160 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 363 | 1 355–1 371 | 5 930 | Apache 2.0 |
| 161 | GPT-5.4 Nano в режиме high | OpenAI | 1 362 | 1 357–1 368 | 19 974 | Proprietary |
| 162 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 362 | 1 355–1 368 | 9 194 | Apache 2.0 |
| 163 | Solar Pro 4 | Upstage | 1 359 | 1 339–1 379 | 873 | Proprietary |
| 164 | R1 | DeepSeek | 1 358 | 1 350–1 365 | 6 426 | MIT |
| 165 | Grok 3 Mini в режиме high | xAI | 1 356 | 1 347–1 366 | 4 127 | Proprietary |
| 166 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 356 | 1 348–1 364 | 6 179 | Apache 2.0 |
| 167 | GPT-5 Mini в режиме high | OpenAI | 1 356 | 1 348–1 363 | 6 905 | Proprietary |
| 168 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 356 | 1 350–1 362 | 12 881 | Proprietary | |
| 169 | GLM 4.6V | Z.ai | 1 354 | 1 332–1 375 | 737 | MIT |
| 170 | GLM 4.5 Air | Z.ai | 1 353 | 1 346–1 360 | 8 062 | MIT |
| 171 | Claude Sonnet 4 · 20250514 | Anthropic | 1 352 | 1 346–1 359 | 9 860 | Proprietary |
| 172 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 352 | 1 346–1 359 | 10 092 | Proprietary |
| 173 | MiniMax M2.5 | MiniMax | 1 352 | 1 346–1 358 | 13 058 | Modified MIT |
| 174 | Nemotron 3 Super | NVIDIA | 1 350 | 1 337–1 363 | 2 009 | NVIDIA Open Model |
| 175 | Kimi K2 0905 | Moonshot AI | 1 346 | 1 335–1 357 | 2 876 | Modified MIT |
| 176 | Granite 4.2 30B | IBM Granite | 1 346 | 1 328–1 365 | 1 104 | Apache 2.0 |
| 177 | Grok 3 Mini | xAI | 1 345 | 1 337–1 353 | 5 400 | Proprietary |
| 178 | DeepSeek V3 0324 | DeepSeek | 1 345 | 1 339–1 351 | 12 339 | MIT |
| 179 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 344 | 1 334–1 354 | 3 492 | Apache 2.0 |
| 180 | Hunyuan TurboS · 20250416 | Tencent | 1 343 | 1 331–1 355 | 2 410 | Proprietary |
| 181 | o1 · preview | OpenAI | 1 342 | 1 335–1 349 | 12 782 | Proprietary |
| 182 | Mistral Medium 3 | Mistral AI | 1 339 | 1 332–1 346 | 7 903 | Proprietary |
| 183 | o3 Mini High | OpenAI | 1 338 | 1 330–1 345 | 6 681 | Proprietary |
| 184 | Gemini 2.0 Flash | 1 336 | 1 330–1 341 | 13 609 | Proprietary | |
| 185 | Trinity Large | Arcee AI | 1 335 | 1 328–1 342 | 9 204 | Apache 2.0 |
| 186 | Qwen2.5 Max | Alibaba Qwen | 1 335 | 1 329–1 341 | 10 937 | Proprietary |
| 187 | Nova 2 Lite | Amazon | 1 334 | 1 324–1 345 | 3 324 | Proprietary |
| 188 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 333 | 1 325–1 341 | 6 159 | Apache 2.0 |
| 189 | Step 3 | StepFun | 1 332 | 1 318–1 347 | 1 631 | Apache 2.0 |
| 190 | MiniMax M2 | MiniMax | 1 332 | 1 319–1 345 | 1 980 | Apache 2.0 |
| 191 | GPT-4.1 Mini | OpenAI | 1 332 | 1 326–1 339 | 9 981 | Proprietary |
| 192 | GLM 4.7 Flash | Z.ai | 1 328 | 1 317–1 338 | 3 143 | MIT |
| 193 | Trinity Large Thinking | Arcee AI | 1 326 | 1 319–1 333 | 9 819 | Apache 2.0 |
| 194 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 325 | 1 319–1 331 | 12 252 | Proprietary |
| 195 | Mercury 2 | Inception Labs | 1 324 | 1 304–1 344 | 839 | Proprietary |
| 196 | Kimi K2 0711 | Moonshot AI | 1 324 | 1 316–1 332 | 6 717 | Modified MIT |
| 197 | Gemma 3 27B | 1 321 | 1 315–1 327 | 12 426 | Gemma | |
| 198 | o4 Mini | OpenAI | 1 321 | 1 315–1 327 | 11 803 | Proprietary |
| 199 | Nemotron 3.5 Lightning | NVIDIA | 1 319 | 1 307–1 331 | 2 888 | OpenMDW-1.1 |
| 200 | gpt-oss-120b | OpenAI | 1 318 | 1 311–1 325 | 7 745 | Apache 2.0 |
| 201 | Ling-flash-2.0 | inclusionAI | 1 318 | 1 304–1 332 | 1 788 | MIT |
| 202 | Hunyuan TurboS · 20250226 | Tencent | 1 316 | 1 298–1 334 | 886 | Proprietary |
| 203 | Amazon Nova Experimental Chat · 10-09 | Amazon | 1 316 | 1 294–1 337 | 705 | Proprietary |
| 204 | o3 Mini | OpenAI | 1 315 | 1 310–1 320 | 16 922 | Proprietary |
| 205 | DeepSeek V3 | DeepSeek | 1 315 | 1 308–1 322 | 8 606 | DeepSeek |
| 206 | INTELLECT-3 | Prime Intellect | 1 313 | 1 297–1 329 | 1 407 | MIT |
| 207 | MiniMax M1 | MiniMax | 1 312 | 1 305–1 319 | 8 545 | Apache 2.0 |
| 208 | GLM 4.5V | Z.ai | 1 312 | 1 296–1 328 | 1 292 | MIT |
| 209 | Ring-flash-2.0 | inclusionAI | 1 312 | 1 298–1 325 | 1 821 | MIT |
| 210 | Mistral Small 3.2 24B | Mistral AI | 1 311 | 1 302–1 319 | 4 360 | Apache 2.0 |
| 211 | Step-1o Turbo | StepFun | 1 310 | 1 298–1 323 | 2 083 | Proprietary |
| 212 | Command A | Cohere | 1 309 | 1 303–1 314 | 15 450 | CC-BY-NC-4.0 |
| 213 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 1 308 | 1 287–1 329 | 660 | Nvidia Open Model |
| 214 | GPT-5 Nano в режиме high | OpenAI | 1 306 | 1 293–1 319 | 2 023 | Proprietary |
| 215 | Qwen3 32B | Alibaba Qwen | 1 305 | 1 286–1 324 | 858 | Apache 2.0 |
| 216 | Gemini 2.0 Flash-Lite | 1 305 | 1 299–1 311 | 9 249 | Proprietary | |
| 217 | o1-mini | OpenAI | 1 304 | 1 299–1 309 | 21 478 | Proprietary |
| 218 | Qwen-Plus | Alibaba Qwen | 1 303 | 1 291–1 315 | 2 249 | Proprietary |
| 219 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 303 | 1 294–1 312 | 4 208 | NVIDIA Open Model |
| 220 | Gemma 3 12B | 1 299 | 1 283–1 315 | 1 145 | Gemma | |
| 221 | OLMo 3.1 32B Instruct | Ai2 | 1 299 | 1 288–1 310 | 3 095 | Apache 2.0 |
| 222 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 298 | 1 279–1 318 | 804 | Nvidia Open |
| 223 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 297 | 1 293–1 302 | 31 212 | Proprietary |
| 224 | QwQ 32B | Alibaba Qwen | 1 297 | 1 290–1 304 | 7 163 | Apache 2.0 |
| 225 | Gemini 1.5 Pro · 1.5-pro-002 | 1 297 | 1 292–1 302 | 22 789 | Proprietary | |
| 226 | Granite 4.2 8B | IBM Granite | 1 294 | 1 275–1 313 | 1 074 | Apache 2.0 |
| 227 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 294 | 1 274–1 313 | 803 | Nvidia |
| 228 | GLM-4-Plus · plus-0111 | Z.ai | 1 293 | 1 280–1 305 | 2 160 | Proprietary |
| 229 | Step-2 16k | StepFun | 1 287 | 1 274–1 299 | 1 950 | Proprietary |
| 230 | Qwen3 30B A3B | Alibaba Qwen | 1 284 | 1 277–1 292 | 6 113 | Apache 2.0 |
| 231 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 280 | 1 269–1 291 | 2 970 | DeepSeek |
| 232 | Hunyuan Turbo | Tencent | 1 279 | 1 261–1 297 | 842 | Proprietary |
| 233 | Yi-Lightning | 01.AI | 1 278 | 1 272–1 285 | 10 932 | Proprietary |
| 234 | GPT-4o (2024-05-13) | OpenAI | 1 278 | 1 273–1 283 | 43 766 | Proprietary |
| 235 | Hunyuan Large | Tencent | 1 277 | 1 262–1 293 | 1 300 | Proprietary |
| 236 | OLMo 3 32B Think | Ai2 | 1 276 | 1 260–1 292 | 1 474 | Apache 2.0 |
| 237 | Molmo 2 8B | Ai2 | 1 275 | 1 237–1 313 | 214 | Apache 2.0 |
| 238 | Qwen2.5 Plus | Alibaba Qwen | 1 275 | 1 266–1 284 | 4 234 | Proprietary |
| 239 | Granite 4.2 3B | IBM Granite | 1 273 | 1 253–1 294 | 1 023 | Apache 2.0 |
| 240 | Athene V2 Chat | Nexusflow | 1 271 | 1 264–1 277 | 10 236 | NexusFlow |
| 241 | Grok 2 | xAI | 1 270 | 1 265–1 275 | 25 659 | Proprietary |
| 242 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 270 | 1 265–1 275 | 32 074 | Proprietary |
| 243 | Llama 4 Maverick | Meta | 1 267 | 1 261–1 274 | 10 470 | Llama 4 |
| 244 | GPT-4o (2024-08-06) | OpenAI | 1 267 | 1 261–1 273 | 18 305 | Proprietary |
| 245 | Granite 4.1 8B | IBM Granite | 1 267 | 1 249–1 284 | 1 390 | Apache 2.0 |
| 246 | GPT-4.1 Nano | OpenAI | 1 267 | 1 254–1 279 | 2 015 | Proprietary |
| 247 | GLM-4-Plus · plus | Z.ai | 1 264 | 1 258–1 271 | 10 743 | Proprietary |
| 248 | Mistral Small 3.1 24B | Mistral AI | 1 264 | 1 257–1 271 | 8 006 | Apache 2.0 |
| 249 | Qwen Max | Alibaba Qwen | 1 262 | 1 254–1 270 | 6 919 | Qwen |
| 250 | Llama 3.1 405B Instruct · bf16 | Meta | 1 259 | 1 254–1 264 | 16 174 | Llama 3.1 Community |
| 251 | Llama 3.1 405B Instruct · fp8 | Meta | 1 259 | 1 254–1 264 | 23 585 | Llama 3.1 Community |
| 252 | GPT-4o-mini (2024-07-18) | OpenAI | 1 259 | 1 254–1 263 | 26 705 | Proprietary |
| 253 | Gemini 1.5 Flash · 002 | 1 258 | 1 252–1 264 | 14 561 | Proprietary | |
| 254 | Gemma 3n E4B | 1 255 | 1 246–1 263 | 4 963 | Gemma | |
| 255 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 254 | 1 249–1 260 | 16 363 | Qwen |
| 256 | Gemini 1.5 Pro · 1.5-pro-001 | 1 254 | 1 248–1 260 | 29 835 | Proprietary | |
| 257 | Gemini 1.5 Pro · advanced-0514 | 1 254 | 1 247–1 260 | 18 524 | Proprietary | |
| 258 | Magistral Medium | Mistral AI | 1 254 | 1 243–1 265 | 3 098 | Proprietary |
| 259 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 252 | 1 241–1 263 | 2 959 | Llama 3.1 |
| 260 | Hunyuan Large Vision | Tencent | 1 250 | 1 233–1 267 | 1 207 | Proprietary |
| 261 | GPT-4 Turbo | OpenAI | 1 250 | 1 244–1 255 | 36 297 | Proprietary |
| 262 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 249 | 1 243–1 256 | 10 175 | DeepSeek |
| 263 | Mistral Large | Mistral AI | 1 249 | 1 243–1 255 | 10 971 | MRL |
| 264 | Claude 3 Opus | Anthropic | 1 248 | 1 244–1 253 | 72 001 | Proprietary |
| 265 | Mistral Large 2407 | Mistral AI | 1 248 | 1 242–1 254 | 18 321 | Mistral Research |
| 266 | Llama 4 Scout | Meta | 1 248 | 1 240–1 255 | 7 448 | Llama |
| 267 | Hunyuan Standard · 2025-02-10 | Tencent | 1 245 | 1 230–1 260 | 1 329 | Proprietary |
| 268 | Grok 2 Mini | xAI | 1 245 | 1 240–1 250 | 21 131 | Proprietary |
| 269 | OLMo 3.1 32B Think | Ai2 | 1 244 | 1 231–1 257 | 2 114 | Apache 2.0 |
| 270 | Llama 3.3 70B Instruct | Meta | 1 242 | 1 237–1 247 | 18 759 | Llama-3.3 |
| 271 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 241 | 1 235–1 246 | 34 416 | Proprietary |
| 272 | Claude 3.5 Haiku | Anthropic | 1 241 | 1 236–1 245 | 21 893 | Proprietary |
| 273 | Mercury | Inception Labs | 1 240 | 1 215–1 266 | 571 | Proprietary |
| 274 | Gemma 3 4B | 1 239 | 1 223–1 255 | 1 233 | Gemma | |
| 275 | gpt-oss-20b | OpenAI | 1 237 | 1 224–1 249 | 2 537 | Apache 2.0 |
| 276 | Nova Pro 1.0 | Amazon | 1 235 | 1 229–1 242 | 9 525 | Proprietary |
| 277 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 235 | 1 229–1 241 | 33 252 | Proprietary |
| 278 | Llama 3.1 Tulu 3 70B | Ai2 | 1 233 | 1 217–1 248 | 1 170 | Llama 3.1 |
| 279 | Llama 3.1 70B Instruct | Meta | 1 231 | 1 226–1 237 | 21 910 | Llama 3.1 Community |
| 280 | Athene 70B | Nexusflow | 1 227 | 1 219–1 235 | 7 490 | CC-BY-NC-4.0 |
| 281 | Qwen2.5 Coder 32B Instruct | Alibaba Qwen | 1 223 | 1 211–1 235 | 2 227 | Apache 2.0 |
| 282 | Granite 4.0 H Small | IBM Granite | 1 218 | 1 203–1 234 | 1 562 | Apache 2.0 |
| 283 | Mistral Small 3 | Mistral AI | 1 214 | 1 206–1 222 | 5 485 | Apache 2.0 |
| 284 | Jamba 1.5 Large | AI21 Labs | 1 213 | 1 202–1 223 | 3 266 | Jamba Open |
| 285 | Gemini 1.5 Flash · 001 | 1 212 | 1 206–1 218 | 23 685 | Proprietary | |
| 286 | Reka Core | Reka AI | 1 212 | 1 202–1 222 | 3 118 | Proprietary |
| 287 | Gemma 2 27B | 1 206 | 1 201–1 211 | 29 545 | Gemma license | |
| 288 | Nova Lite 1.0 | Amazon | 1 205 | 1 198–1 212 | 7 809 | Proprietary |
| 289 | Nemotron-4 340B Instruct | NVIDIA | 1 203 | 1 195–1 211 | 7 354 | NVIDIA Open Model |
| 290 | GLM-4 | Z.ai | 1 203 | 1 193–1 213 | 3 766 | Proprietary |
| 291 | Hunyuan Standard · 256k | Tencent | 1 202 | 1 185–1 219 | 1 098 | Proprietary |
| 292 | Phi 4 | Microsoft | 1 201 | 1 195–1 208 | 9 162 | MIT |
| 293 | GPT-4 | OpenAI | 1 201 | 1 194–1 208 | 18 087 | Proprietary |
| 294 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 201 | 1 186–1 215 | 1 484 | Llama 3.1 |
| 295 | Claude 3 Sonnet | Anthropic | 1 200 | 1 194–1 205 | 38 802 | Proprietary |
| 296 | Gemini 1.5 Flash-8B | 1 199 | 1 193–1 205 | 14 894 | Proprietary | |
| 297 | Command R+ (08-2024) | Cohere | 1 197 | 1 188–1 206 | 4 024 | CC-BY-NC-4.0 |
| 298 | Aya Expanse 32B | Cohere | 1 196 | 1 190–1 203 | 11 265 | CC-BY-NC-4.0 |
| 299 | Llama 3 70B Instruct | Meta | 1 194 | 1 189–1 200 | 56 558 | Llama 3 Community |
| 300 | Gemma 2 9B IT SimPO | Princeton NLP | 1 191 | 1 181–1 201 | 3 741 | MIT |
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.