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: тексты, литература, языки», 300 строк. Оценка и интервал — из источника; интервал не равен диапазону мест.
| Место | Модель | Разработчик | Оценка | Интервал | Голоса | Лицензия |
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5.1 в режиме max | Anthropic | 1 516 | 1 500–1 532 | 1 574 | Proprietary |
| 2 | Claude Fable 5 | Anthropic | 1 504 | 1 497–1 512 | 8 123 | Proprietary |
| 3 | Gemini 3.8 Flash в режиме high | 1 502 | 1 485–1 518 | 1 378 | Proprietary | |
| 4 | Claude Opus 4.6 в режиме high | Anthropic | 1 502 | 1 496–1 507 | 18 156 | Proprietary |
| 5 | Claude Opus 5 в режиме max | Anthropic | 1 501 | 1 493–1 510 | 5 802 | Proprietary |
| 6 | Claude Opus 5 в режиме high | Anthropic | 1 499 | 1 492–1 506 | 11 811 | Proprietary |
| 7 | Claude Opus 4.6 | Anthropic | 1 494 | 1 488–1 499 | 18 346 | Proprietary |
| 8 | Claude Opus 4.7 в режиме high | Anthropic | 1 490 | 1 484–1 497 | 14 966 | Proprietary |
| 9 | Gemini 3.7 Flash в режиме high | 1 488 | 1 472–1 504 | 1 480 | Proprietary | |
| 10 | Gemini 3.1 Pro Preview | 1 480 | 1 475–1 485 | 26 539 | Proprietary | |
| 11 | Claude Opus 4.7 | Anthropic | 1 479 | 1 473–1 485 | 15 324 | Proprietary |
| 12 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 478 | 1 469–1 488 | 4 238 | Proprietary |
| 13 | Gemini 3 Pro | 1 478 | 1 471–1 485 | 9 182 | Proprietary | |
| 14 | Gemini 3.5 Flash в режиме high | 1 478 | 1 471–1 484 | 9 743 | Proprietary | |
| 15 | GLM 5.3 в режиме max | Z.ai | 1 473 | 1 462–1 484 | 3 095 | MIT |
| 16 | Gemini 3.5 Flash в режиме medium | 1 472 | 1 465–1 479 | 9 678 | Proprietary | |
| 17 | Gemini 3.6 Flash в режиме high | 1 471 | 1 463–1 479 | 7 037 | Proprietary | |
| 18 | Muse Spark 1.2 в режиме xhigh | Meta | 1 470 | 1 450–1 491 | 845 | Proprietary |
| 19 | Qwen3.5 Max | Alibaba Qwen | 1 467 | 1 459–1 476 | 4 896 | Proprietary |
| 20 | GPT-5.5 в режиме high | OpenAI | 1 467 | 1 461–1 473 | 16 595 | Proprietary |
| 21 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 465 | 1 457–1 473 | 7 052 | Proprietary |
| 22 | Kimi K3 в режиме max | Moonshot AI | 1 464 | 1 455–1 473 | 5 445 | Kimi K3 license |
| 23 | GPT-5.5 · 5.5 | OpenAI | 1 462 | 1 456–1 468 | 16 943 | Proprietary |
| 24 | Muse Spark 1.3 в режиме max | Meta | 1 462 | 1 444–1 479 | 1 210 | Proprietary |
| 25 | Claude Opus 4.8 в режиме high | Anthropic | 1 460 | 1 454–1 467 | 13 740 | Proprietary |
| 26 | Gemini 3 Flash Preview | 1 459 | 1 451–1 467 | 6 801 | Proprietary | |
| 27 | GPT-5.4 в режиме high | OpenAI | 1 459 | 1 453–1 465 | 15 014 | Proprietary |
| 28 | GLM 5.2 в режиме max | Z.ai | 1 458 | 1 451–1 465 | 9 635 | MIT |
| 29 | Qwen3.7 Max | Alibaba Qwen | 1 457 | 1 436–1 478 | 851 | Proprietary |
| 30 | GLM 5.1 | Z.ai | 1 455 | 1 448–1 461 | 12 603 | MIT |
| 31 | Muse Spark 1.1 | Meta | 1 455 | 1 447–1 462 | 7 307 | Proprietary |
| 32 | Gemini 2.5 Pro | 1 453 | 1 449–1 458 | 27 475 | Proprietary | |
| 33 | MiMo-V2.5-Pro | Xiaomi | 1 453 | 1 447–1 459 | 15 482 | MIT |
| 34 | Claude Opus 4.8 | Anthropic | 1 452 | 1 446–1 459 | 13 777 | Proprietary |
| 35 | Muse Spark | Meta | 1 452 | 1 441–1 463 | 3 130 | Proprietary |
| 36 | Claude Opus 4.5 | Anthropic | 1 450 | 1 445–1 456 | 16 156 | Proprietary |
| 37 | GPT-5.4 | OpenAI | 1 450 | 1 444–1 456 | 15 360 | Proprietary |
| 38 | GLM 5.3 Flash | Z.ai | 1 449 | 1 438–1 461 | 2 753 | MIT |
| 39 | GPT-6 Astra в режиме max | OpenAI | 1 449 | 1 426–1 472 | 714 | Proprietary |
| 40 | ERNIE 5.1 | Baidu | 1 448 | 1 441–1 456 | 8 948 | Proprietary |
| 41 | DeepSeek V4 Pro 0423 | DeepSeek | 1 446 | 1 440–1 453 | 13 127 | MIT |
| 42 | Grok 4.5 | xAI | 1 446 | 1 438–1 453 | 7 962 | Proprietary |
| 43 | Qwen3.6 Max Preview | Alibaba Qwen | 1 446 | 1 427–1 464 | 1 119 | Proprietary |
| 44 | Claude Sonnet 4.6 | Anthropic | 1 445 | 1 439–1 451 | 16 226 | Proprietary |
| 45 | Claude Opus 4.5 в режиме high | Anthropic | 1 445 | 1 438–1 452 | 8 251 | Proprietary |
| 46 | Claude Sonnet 4.5 | Anthropic | 1 442 | 1 437–1 447 | 18 218 | Proprietary |
| 47 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 441 | 1 429–1 453 | 2 411 | MIT |
| 48 | Qwen3.7 Plus | Alibaba Qwen | 1 439 | 1 432–1 446 | 9 962 | Proprietary |
| 49 | GLM 5 | Z.ai | 1 438 | 1 431–1 446 | 6 550 | MIT |
| 50 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 437 | 1 430–1 443 | 12 724 | MIT |
| 51 | Kimi K2.6 | Moonshot AI | 1 437 | 1 429–1 444 | 8 851 | Modified MIT |
| 52 | Gemini 3 Flash Preview в режиме minimal | 1 436 | 1 431–1 442 | 20 426 | Proprietary | |
| 53 | Grok 4.20 · beta-0309-reasoning | xAI | 1 436 | 1 430–1 442 | 15 266 | Proprietary |
| 54 | Claude Sonnet 5 в режиме high | Anthropic | 1 435 | 1 428–1 442 | 9 319 | Proprietary |
| 55 | Grok 4.6 в режиме high | xAI | 1 432 | 1 422–1 442 | 4 250 | Proprietary |
| 56 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 432 | 1 427–1 437 | 18 629 | Proprietary |
| 57 | Grok 4.20 Multi-Agent | xAI | 1 432 | 1 426–1 438 | 14 754 | Proprietary |
| 58 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 431 | 1 423–1 439 | 7 544 | Proprietary |
| 59 | Gemma 4 31B | 1 429 | 1 413–1 445 | 1 355 | Apache 2.0 | |
| 60 | Gemini 3.5 Flash Lite | 1 428 | 1 419–1 436 | 7 002 | Proprietary | |
| 61 | Grok 4.20 · beta1 | xAI | 1 427 | 1 419–1 435 | 6 068 | Proprietary |
| 62 | GPT-5.1 в режиме high | OpenAI | 1 426 | 1 420–1 433 | 9 006 | Proprietary |
| 63 | Kimi K2.5 · thinking | Moonshot AI | 1 426 | 1 421–1 432 | 16 639 | Modified MIT |
| 64 | MiMo-V2 Pro | Xiaomi | 1 424 | 1 416–1 433 | 5 636 | Proprietary |
| 65 | GPT-5.5 · 5.5-instant | OpenAI | 1 424 | 1 416–1 432 | 6 294 | Proprietary |
| 66 | GPT-5.2 Chat | OpenAI | 1 421 | 1 414–1 429 | 8 047 | Proprietary |
| 67 | ERNIE 5.0 · 0110 | Baidu | 1 421 | 1 414–1 428 | 7 941 | Proprietary |
| 68 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 420 | 1 412–1 427 | 7 589 | Proprietary |
| 69 | Qwen3.6 Plus | Alibaba Qwen | 1 419 | 1 412–1 426 | 10 480 | Proprietary |
| 70 | MiniMax M3 | MiniMax | 1 418 | 1 411–1 424 | 12 049 | MiniMax Community License |
| 71 | Claude Opus 4.1 · 20250805 | Anthropic | 1 417 | 1 412–1 423 | 17 072 | Proprietary |
| 72 | GLM 5V Turbo | Z.ai | 1 417 | 1 404–1 430 | 2 388 | Proprietary |
| 73 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 417 | 1 410–1 423 | 10 857 | Proprietary |
| 74 | Qwen3.5 397B A17B | Alibaba Qwen | 1 416 | 1 411–1 421 | 18 410 | Apache 2.0 |
| 75 | ChatGPT-4o (latest) | OpenAI | 1 415 | 1 410–1 420 | 18 277 | Proprietary |
| 76 | GLM 4.6 | Z.ai | 1 415 | 1 408–1 422 | 7 875 | MIT |
| 77 | ERNIE 5.0 · preview-1203 | Baidu | 1 414 | 1 401–1 426 | 2 133 | Proprietary |
| 78 | GPT-5.1 | OpenAI | 1 412 | 1 406–1 419 | 9 734 | Proprietary |
| 79 | DeepSeek V4 Flash 0423 | DeepSeek | 1 412 | 1 406–1 419 | 11 734 | MIT |
| 80 | DeepSeek V3.1 Terminus · terminus-thinking | DeepSeek | 1 412 | 1 391–1 433 | 754 | MIT |
| 81 | GPT-4.5 Preview | OpenAI | 1 411 | 1 402–1 421 | 4 092 | Proprietary |
| 82 | Grok 3 | xAI | 1 411 | 1 404–1 418 | 7 725 | Proprietary |
| 83 | Hy3 | Tencent | 1 410 | 1 397–1 424 | 2 086 | Apache 2.0 |
| 84 | GLM 4.7 | Z.ai | 1 410 | 1 399–1 421 | 2 734 | MIT |
| 85 | ERNIE 5.0 · preview-1022 | Baidu | 1 410 | 1 393–1 427 | 1 112 | Proprietary |
| 86 | Gemma 4 26B A4B | 1 410 | 1 394–1 425 | 1 324 | Apache 2.0 | |
| 87 | Seed 2.0 Pro | ByteDance Seed | 1 409 | 1 403–1 414 | 17 634 | Proprietary |
| 88 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 408 | 1 397–1 420 | 2 595 | MIT |
| 89 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 408 | 1 401–1 415 | 11 827 | MIT |
| 90 | MiMo-V2.5 | Xiaomi | 1 408 | 1 401–1 415 | 10 758 | MIT |
| 91 | Gemini 3.1 Flash Lite Preview | 1 408 | 1 402–1 414 | 14 480 | Proprietary | |
| 92 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 408 | 1 401–1 414 | 10 300 | MIT |
| 93 | Grok 4.1 · 4.1 | xAI | 1 407 | 1 401–1 413 | 14 746 | Proprietary |
| 94 | DeepSeek V3.1 Terminus · terminus | DeepSeek | 1 407 | 1 386–1 428 | 788 | MIT |
| 95 | Gemini 2.5 Flash · flash | 1 407 | 1 402–1 411 | 27 771 | Proprietary | |
| 96 | Qwen3 Max · preview | Alibaba Qwen | 1 406 | 1 399–1 414 | 6 131 | Proprietary |
| 97 | Grok 4.1 · 4.1-thinking | xAI | 1 405 | 1 400–1 411 | 14 658 | Proprietary |
| 98 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 404 | 1 393–1 416 | 2 720 | MIT |
| 99 | GLM 4.5 | Z.ai | 1 403 | 1 395–1 411 | 5 272 | MIT |
| 100 | R1 0528 | DeepSeek | 1 402 | 1 392–1 411 | 4 170 | MIT |
| 101 | Qwen3.8 27B | Alibaba Qwen | 1 401 | 1 389–1 413 | 2 694 | Apache 2.0 |
| 102 | Nemotron 3 Ultra | NVIDIA | 1 401 | 1 389–1 413 | 2 737 | OpenMDW-1.1 |
| 103 | Inkling | Thinking Machines Lab | 1 401 | 1 393–1 409 | 6 836 | Apache 2.0 |
| 104 | Grok 4 | xAI | 1 399 | 1 393–1 406 | 8 868 | Proprietary |
| 105 | Kimi K2.5 · instant | Moonshot AI | 1 398 | 1 385–1 412 | 1 828 | Modified MIT |
| 106 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 398 | 1 391–1 404 | 9 350 | MIT |
| 107 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 397 | 1 387–1 408 | 3 423 | MIT |
| 108 | Mistral Large 3 2512 | Mistral AI | 1 397 | 1 391–1 402 | 16 657 | Apache 2.0 |
| 109 | Mistral Medium 3.1 | Mistral AI | 1 396 | 1 391–1 400 | 21 216 | Proprietary |
| 110 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 396 | 1 383–1 409 | 2 082 | MIT |
| 111 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 395 | 1 388–1 403 | 7 209 | Proprietary | |
| 112 | MiMo-V2 Omni | Xiaomi | 1 395 | 1 385–1 404 | 4 812 | Proprietary |
| 113 | GPT-5.4 Mini в режиме high | OpenAI | 1 394 | 1 388–1 400 | 14 530 | Proprietary |
| 114 | GPT-5.2 в режиме high | OpenAI | 1 392 | 1 386–1 399 | 10 988 | Proprietary |
| 115 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 392 | 1 379–1 405 | 2 320 | Apache 2.0 |
| 116 | GPT-5.2 | OpenAI | 1 392 | 1 387–1 397 | 18 982 | Proprietary |
| 117 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 391 | 1 384–1 399 | 7 771 | Proprietary |
| 118 | Grok 4 Fast · chat | xAI | 1 390 | 1 375–1 406 | 1 456 | Proprietary |
| 119 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 390 | 1 386–1 395 | 21 523 | Apache 2.0 |
| 120 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 389 | 1 376–1 402 | 1 949 | Proprietary |
| 121 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 389 | 1 381–1 397 | 6 496 | Proprietary |
| 122 | Amazon Nova Experimental Chat · 26-02-10 | Amazon | 1 388 | 1 367–1 409 | 688 | Proprietary |
| 123 | Hunyuan Vision 1.5 | Tencent | 1 388 | 1 363–1 414 | 491 | Proprietary |
| 124 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 388 | 1 380–1 396 | 6 454 | Apache 2.0 |
| 125 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 388 | 1 374–1 401 | 1 849 | Apache 2.0 |
| 126 | Grok 4.1 Fast | xAI | 1 387 | 1 381–1 393 | 12 838 | Proprietary |
| 127 | Claude Haiku 4.5 | Anthropic | 1 387 | 1 383–1 391 | 30 598 | Proprietary |
| 128 | Grok 4.3 | xAI | 1 386 | 1 381–1 392 | 17 054 | Proprietary |
| 129 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 385 | 1 379–1 391 | 10 434 | MIT |
| 130 | GPT-5 | OpenAI | 1 385 | 1 377–1 392 | 6 706 | Proprietary |
| 131 | Hunyuan T1 | Tencent | 1 384 | 1 366–1 403 | 1 008 | Proprietary |
| 132 | Mistral Medium 3.5 | Mistral AI | 1 384 | 1 372–1 396 | 2 722 | Modified MIT |
| 133 | Claude Opus 4 · 20250514 | Anthropic | 1 381 | 1 374–1 388 | 9 256 | Proprietary |
| 134 | Amazon Nova Experimental Chat · 12-10 | Amazon | 1 381 | 1 361–1 401 | 827 | Proprietary |
| 135 | Kimi K2 Thinking | Moonshot AI | 1 380 | 1 375–1 386 | 13 997 | Modified MIT |
| 136 | Qwen3.5-27B | Alibaba Qwen | 1 379 | 1 372–1 387 | 6 144 | Apache 2.0 |
| 137 | GPT-5 в режиме high | OpenAI | 1 379 | 1 372–1 387 | 6 962 | Proprietary |
| 138 | Grok 4 Fast · reasoning | xAI | 1 379 | 1 370–1 388 | 4 095 | Proprietary |
| 139 | GPT-5.3 Chat | OpenAI | 1 379 | 1 371–1 386 | 7 597 | Proprietary |
| 140 | Hy3 preview | Tencent | 1 375 | 1 359–1 391 | 1 522 | tencent-hunyuan-community |
| 141 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 374 | 1 367–1 382 | 7 031 | Proprietary | |
| 142 | MiniMax M2.7 | MiniMax | 1 373 | 1 368–1 379 | 16 467 | Modified MIT |
| 143 | MiniMax M2.1 | MiniMax | 1 373 | 1 363–1 382 | 3 884 | MIT |
| 144 | Step 3.5 Flash | StepFun | 1 372 | 1 366–1 378 | 13 910 | Apache 2.0 |
| 145 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 372 | 1 364–1 379 | 5 642 | Proprietary |
| 146 | o3 | OpenAI | 1 371 | 1 365–1 377 | 13 151 | Proprietary |
| 147 | GPT-4.1 | OpenAI | 1 371 | 1 365–1 377 | 11 233 | Proprietary |
| 148 | Qwen3.5-Flash | Alibaba Qwen | 1 369 | 1 363–1 375 | 13 598 | Proprietary |
| 149 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 368 | 1 362–1 374 | 10 360 | Proprietary | |
| 150 | LongCat-Flash-Chat · chat | Meituan | 1 367 | 1 355–1 379 | 2 473 | MIT |
| 151 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 366 | 1 359–1 374 | 7 982 | Apache 2.0 |
| 152 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 366 | 1 358–1 373 | 6 669 | Apache 2.0 |
| 153 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 366 | 1 352–1 380 | 1 737 | Apache 2.0 |
| 154 | DeepSeek V3 0324 | DeepSeek | 1 365 | 1 358–1 371 | 10 248 | MIT |
| 155 | R1 | DeepSeek | 1 364 | 1 356–1 372 | 5 355 | MIT |
| 156 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 364 | 1 352–1 376 | 2 399 | MIT |
| 157 | o1 · 2024-12-17 | OpenAI | 1 360 | 1 353–1 367 | 7 663 | Proprietary |
| 158 | Muse Glimmer 30B | Meta | 1 359 | 1 339–1 380 | 905 | Apache-2.0 |
| 159 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 356 | 1 349–1 364 | 7 323 | Proprietary |
| 160 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 356 | 1 347–1 364 | 5 008 | Apache 2.0 |
| 161 | GLM 4.5 Air | Z.ai | 1 353 | 1 346–1 361 | 6 895 | MIT |
| 162 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 352 | 1 332–1 373 | 754 | Proprietary |
| 163 | Mistral Medium 3 | Mistral AI | 1 351 | 1 344–1 359 | 7 144 | Proprietary |
| 164 | Hunyuan TurboS · 20250416 | Tencent | 1 351 | 1 339–1 363 | 2 514 | Proprietary |
| 165 | Claude Sonnet 4 · 20250514 | Anthropic | 1 351 | 1 344–1 358 | 8 451 | Proprietary |
| 166 | Inkling Small | Thinking Machines Lab | 1 350 | 1 340–1 359 | 5 112 | Apache 2.0 |
| 167 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 349 | 1 337–1 361 | 2 501 | Proprietary |
| 168 | GLM 4.6V | Z.ai | 1 348 | 1 325–1 372 | 600 | MIT |
| 169 | MiniMax M2.5 | MiniMax | 1 348 | 1 341–1 354 | 9 601 | Modified MIT |
| 170 | Gemma 3 27B | 1 347 | 1 341–1 353 | 10 542 | Gemma | |
| 171 | Gemini 2.0 Flash | 1 347 | 1 341–1 353 | 10 517 | Proprietary | |
| 172 | Qwen2.5 Max | Alibaba Qwen | 1 346 | 1 339–1 353 | 8 174 | Proprietary |
| 173 | Grok 3 Mini в режиме high | xAI | 1 345 | 1 335–1 355 | 3 408 | Proprietary |
| 174 | Grok 3 Mini | xAI | 1 345 | 1 336–1 353 | 4 531 | Proprietary |
| 175 | GPT-5 Mini в режиме high | OpenAI | 1 344 | 1 337–1 352 | 6 079 | Proprietary |
| 176 | Kimi K2 0905 | Moonshot AI | 1 344 | 1 332–1 356 | 2 576 | Modified MIT |
| 177 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 340 | 1 331–1 348 | 5 293 | Apache 2.0 |
| 178 | GPT-5.4 Nano в режиме high | OpenAI | 1 339 | 1 333–1 346 | 14 198 | Proprietary |
| 179 | o1 · preview | OpenAI | 1 339 | 1 331–1 347 | 8 360 | Proprietary |
| 180 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 338 | 1 331–1 345 | 8 741 | Proprietary |
| 181 | INTELLECT-3 | Prime Intellect | 1 338 | 1 321–1 354 | 1 266 | MIT |
| 182 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 336 | 1 328–1 344 | 6 116 | Apache 2.0 |
| 183 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 336 | 1 327–1 344 | 5 038 | Apache 2.0 |
| 184 | Nemotron 3 Super | NVIDIA | 1 331 | 1 317–1 346 | 1 619 | NVIDIA Open Model |
| 185 | DeepSeek V3 | DeepSeek | 1 331 | 1 323–1 339 | 5 913 | DeepSeek |
| 186 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 330 | 1 319–1 340 | 3 039 | Apache 2.0 |
| 187 | Trinity Large | Arcee AI | 1 329 | 1 322–1 337 | 6 855 | Apache 2.0 |
| 188 | Kimi K2 0711 | Moonshot AI | 1 328 | 1 320–1 336 | 5 943 | Modified MIT |
| 189 | Command A | Cohere | 1 327 | 1 321–1 333 | 12 705 | CC-BY-NC-4.0 |
| 190 | GLM-4-Plus · plus-0111 | Z.ai | 1 326 | 1 311–1 341 | 1 459 | Proprietary |
| 191 | Gemini 1.5 Pro · 1.5-pro-002 | 1 326 | 1 320–1 332 | 14 953 | Proprietary | |
| 192 | Step 3 | StepFun | 1 326 | 1 310–1 341 | 1 378 | Apache 2.0 |
| 193 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 324 | 1 317–1 331 | 9 675 | Proprietary |
| 194 | Solar Pro 4 | Upstage | 1 323 | 1 299–1 347 | 661 | Proprietary |
| 195 | Gemini 2.0 Flash-Lite | 1 321 | 1 314–1 329 | 6 681 | Proprietary | |
| 196 | GLM 4.7 Flash | Z.ai | 1 321 | 1 309–1 333 | 2 516 | MIT |
| 197 | Amazon Nova Experimental Chat · 10-09 | Amazon | 1 321 | 1 298–1 344 | 611 | Proprietary |
| 198 | Trinity Large Thinking | Arcee AI | 1 318 | 1 310–1 326 | 6 900 | Apache 2.0 |
| 199 | GPT-4.1 Mini | OpenAI | 1 318 | 1 311–1 325 | 8 624 | Proprietary |
| 200 | Nova 2 Lite | Amazon | 1 318 | 1 307–1 329 | 2 740 | Proprietary |
| 201 | o4 Mini | OpenAI | 1 316 | 1 309–1 322 | 9 967 | Proprietary |
| 202 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 313 | 1 308–1 318 | 21 785 | Proprietary |
| 203 | o3 Mini High | OpenAI | 1 313 | 1 304–1 322 | 4 762 | Proprietary |
| 204 | MiniMax M1 | MiniMax | 1 311 | 1 304–1 318 | 7 635 | Apache 2.0 |
| 205 | Mistral Small 3.2 24B | Mistral AI | 1 310 | 1 300–1 320 | 3 740 | Apache 2.0 |
| 206 | Step-1o Turbo | StepFun | 1 309 | 1 296–1 323 | 1 819 | Proprietary |
| 207 | Granite 4.2 30B | IBM Granite | 1 307 | 1 286–1 329 | 824 | Apache 2.0 |
| 208 | Gemma 3 12B | 1 307 | 1 290–1 324 | 1 010 | Gemma | |
| 209 | gpt-oss-120b | OpenAI | 1 305 | 1 297–1 313 | 6 586 | Apache 2.0 |
| 210 | Mercury 2 | Inception Labs | 1 304 | 1 282–1 326 | 686 | Proprietary |
| 211 | Step-2 16k | StepFun | 1 303 | 1 288–1 319 | 1 326 | Proprietary |
| 212 | Hunyuan TurboS · 20250226 | Tencent | 1 302 | 1 281–1 323 | 662 | Proprietary |
| 213 | GLM 4.5V | Z.ai | 1 302 | 1 284–1 319 | 1 067 | MIT |
| 214 | QwQ 32B | Alibaba Qwen | 1 301 | 1 293–1 309 | 6 102 | Apache 2.0 |
| 215 | Llama 3.3 Nemotron Super 49B v1 | NVIDIA | 1 300 | 1 278–1 323 | 614 | Nvidia |
| 216 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 1 300 | 1 279–1 321 | 712 | Nvidia Open Model |
| 217 | Qwen3 32B | Alibaba Qwen | 1 298 | 1 282–1 315 | 1 121 | Apache 2.0 |
| 218 | GPT-4o (2024-05-13) | OpenAI | 1 298 | 1 292–1 303 | 29 959 | Proprietary |
| 219 | MiniMax M2 | MiniMax | 1 297 | 1 282–1 313 | 1 459 | Apache 2.0 |
| 220 | Ring-flash-2.0 | inclusionAI | 1 296 | 1 281–1 312 | 1 460 | MIT |
| 221 | Qwen-Plus | Alibaba Qwen | 1 296 | 1 281–1 311 | 1 512 | Proprietary |
| 222 | o3 Mini | OpenAI | 1 294 | 1 289–1 300 | 13 377 | Proprietary |
| 223 | Llama 3.3 Nemotron Super 49B v1.5 | NVIDIA | 1 294 | 1 272–1 316 | 661 | Nvidia Open |
| 224 | Ling-flash-2.0 | inclusionAI | 1 293 | 1 278–1 309 | 1 441 | MIT |
| 225 | Gemini 1.5 Flash · 002 | 1 293 | 1 286–1 300 | 9 441 | Proprietary | |
| 226 | Grok 2 | xAI | 1 292 | 1 286–1 298 | 17 406 | Proprietary |
| 227 | Qwen3 30B A3B | Alibaba Qwen | 1 290 | 1 283–1 298 | 5 918 | Apache 2.0 |
| 228 | Hunyuan Turbo | Tencent | 1 290 | 1 271–1 310 | 711 | Proprietary |
| 229 | GPT-4o (2024-08-06) | OpenAI | 1 289 | 1 283–1 296 | 12 147 | Proprietary |
| 230 | Gemini 1.5 Pro · 1.5-pro-001 | 1 289 | 1 282–1 295 | 20 950 | Proprietary | |
| 231 | Gemma 3 4B | 1 288 | 1 271–1 305 | 1 080 | Gemma | |
| 232 | Gemma 3n E4B | 1 287 | 1 278–1 295 | 4 850 | Gemma | |
| 233 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 286 | 1 276–1 296 | 3 455 | NVIDIA Open Model |
| 234 | Gemini 1.5 Pro · advanced-0514 | 1 286 | 1 278–1 294 | 13 201 | Proprietary | |
| 235 | GPT-5 Nano в режиме high | OpenAI | 1 284 | 1 270–1 298 | 1 724 | Proprietary |
| 236 | Nemotron 3.5 Lightning | NVIDIA | 1 284 | 1 270–1 298 | 2 211 | OpenMDW-1.1 |
| 237 | DeepSeek V2.5 · v2.5-1210 | DeepSeek | 1 283 | 1 270–1 296 | 1 930 | DeepSeek |
| 238 | GPT-4o-mini (2024-07-18) | OpenAI | 1 282 | 1 276–1 287 | 18 531 | Proprietary |
| 239 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 280 | 1 274–1 286 | 21 916 | Proprietary |
| 240 | o1-mini | OpenAI | 1 280 | 1 274–1 286 | 14 272 | Proprietary |
| 241 | GPT-4 Turbo | OpenAI | 1 278 | 1 272–1 285 | 24 529 | Proprietary |
| 242 | GLM-4-Plus · plus | Z.ai | 1 278 | 1 270–1 287 | 7 164 | Proprietary |
| 243 | Llama 4 Maverick | Meta | 1 277 | 1 270–1 284 | 8 705 | Llama 4 |
| 244 | OLMo 3.1 32B Instruct | Ai2 | 1 276 | 1 264–1 288 | 2 545 | Apache 2.0 |
| 245 | Yi-Lightning | 01.AI | 1 276 | 1 268–1 284 | 7 632 | Proprietary |
| 246 | Llama 3.1 405B Instruct · fp8 | Meta | 1 274 | 1 268–1 280 | 15 999 | Llama 3.1 Community |
| 247 | Qwen2.5 Plus | Alibaba Qwen | 1 273 | 1 262–1 284 | 2 823 | Proprietary |
| 248 | Granite 4.2 8B | IBM Granite | 1 272 | 1 248–1 295 | 775 | Apache 2.0 |
| 249 | Hunyuan Large | Tencent | 1 271 | 1 252–1 289 | 890 | Proprietary |
| 250 | Llama 3.1 405B Instruct · bf16 | Meta | 1 270 | 1 264–1 276 | 11 150 | Llama 3.1 Community |
| 251 | Qwen Max | Alibaba Qwen | 1 270 | 1 261–1 279 | 4 631 | Qwen |
| 252 | OLMo 3 32B Think | Ai2 | 1 270 | 1 253–1 287 | 1 302 | Apache 2.0 |
| 253 | Llama 4 Scout | Meta | 1 269 | 1 261–1 277 | 6 679 | Llama |
| 254 | Claude 3 Opus | Anthropic | 1 268 | 1 263–1 274 | 49 929 | Proprietary |
| 255 | Mistral Small 3.1 24B | Mistral AI | 1 268 | 1 261–1 276 | 7 039 | Apache 2.0 |
| 256 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 264 | 1 258–1 271 | 24 682 | Proprietary |
| 257 | Grok 2 Mini | xAI | 1 262 | 1 256–1 268 | 14 508 | Proprietary |
| 258 | Hunyuan Large Vision | Tencent | 1 261 | 1 243–1 279 | 1 077 | Proprietary |
| 259 | GPT-4.1 Nano | OpenAI | 1 261 | 1 247–1 275 | 1 583 | Proprietary |
| 260 | Mistral Large | Mistral AI | 1 260 | 1 253–1 267 | 7 416 | MRL |
| 261 | Granite 4.1 8B | IBM Granite | 1 260 | 1 239–1 281 | 948 | Apache 2.0 |
| 262 | Mistral Large 2407 | Mistral AI | 1 258 | 1 251–1 264 | 12 224 | Mistral Research |
| 263 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 257 | 1 250–1 264 | 23 492 | Proprietary |
| 264 | Hunyuan Standard | Tencent | 1 257 | 1 239–1 275 | 937 | Proprietary |
| 265 | Magistral Medium | Mistral AI | 1 257 | 1 244–1 269 | 2 385 | Proprietary |
| 266 | Claude 3.5 Haiku | Anthropic | 1 256 | 1 251–1 261 | 16 795 | Proprietary |
| 267 | Llama 3.3 70B Instruct | Meta | 1 255 | 1 249–1 260 | 13 783 | Llama-3.3 |
| 268 | Llama 3.1 Tulu 3 70B | Ai2 | 1 254 | 1 235–1 273 | 758 | Llama 3.1 |
| 269 | Athene V2 Chat | Nexusflow | 1 254 | 1 246–1 262 | 6 686 | NexusFlow |
| 270 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 1 252 | 1 239–1 266 | 1 828 | Llama 3.1 |
| 271 | Gemini 1.5 Flash · 001 | 1 252 | 1 245–1 258 | 16 463 | Proprietary | |
| 272 | OLMo 3.1 32B Think | Ai2 | 1 250 | 1 236–1 264 | 1 883 | Apache 2.0 |
| 273 | Gemma 2 27B | 1 249 | 1 243–1 254 | 20 390 | Gemma license | |
| 274 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 247 | 1 240–1 253 | 10 827 | Qwen |
| 275 | DeepSeek V2.5 · v2.5 | DeepSeek | 1 247 | 1 238–1 255 | 6 613 | DeepSeek |
| 276 | Nova Pro 1.0 | Amazon | 1 246 | 1 239–1 254 | 6 609 | Proprietary |
| 277 | Reka Core | Reka AI | 1 242 | 1 229–1 256 | 1 890 | Proprietary |
| 278 | Command R+ (08-2024) | Cohere | 1 241 | 1 230–1 252 | 2 672 | CC-BY-NC-4.0 |
| 279 | Athene 70B | Nexusflow | 1 241 | 1 232–1 249 | 5 378 | CC-BY-NC-4.0 |
| 280 | gpt-oss-20b | OpenAI | 1 237 | 1 224–1 250 | 2 166 | Apache 2.0 |
| 281 | Llama 3.1 70B Instruct | Meta | 1 235 | 1 229–1 242 | 14 962 | Llama 3.1 Community |
| 282 | Mercury | Inception Labs | 1 228 | 1 199–1 257 | 457 | Proprietary |
| 283 | Jamba 1.5 Large | AI21 Labs | 1 227 | 1 215–1 240 | 2 350 | Jamba Open |
| 284 | Granite 4.2 3B | IBM Granite | 1 227 | 1 202–1 252 | 746 | Apache 2.0 |
| 285 | Gemini 1.5 Flash-8B | 1 227 | 1 219–1 234 | 9 499 | Proprietary | |
| 286 | Gemma 2 9B IT SimPO | Princeton NLP | 1 226 | 1 214–1 237 | 2 801 | MIT |
| 287 | Granite 4.0 H Small | IBM Granite | 1 224 | 1 206–1 242 | 1 175 | Apache 2.0 |
| 288 | Nemotron-4 340B Instruct | NVIDIA | 1 223 | 1 214–1 232 | 5 386 | NVIDIA Open Model |
| 289 | Aya Expanse 32B | Cohere | 1 222 | 1 214–1 230 | 7 297 | CC-BY-NC-4.0 |
| 290 | Llama 3.1 Nemotron 51B Instruct | NVIDIA | 1 220 | 1 202–1 238 | 1 007 | Llama 3.1 |
| 291 | Gemma 2 9B | 1 219 | 1 213–1 225 | 14 640 | Gemma license | |
| 292 | GPT-4 · 0314 | OpenAI | 1 218 | 1 210–1 227 | 12 861 | Proprietary |
| 293 | Command R+ | Cohere | 1 218 | 1 211–1 226 | 19 567 | CC-BY-NC-4.0 |
| 294 | Claude 3 Sonnet | Anthropic | 1 217 | 1 211–1 224 | 27 297 | Proprietary |
| 295 | Reka Flash (2024-09) | Reka AI | 1 217 | 1 204–1 230 | 1 950 | Proprietary |
| 296 | Mistral Small 3 | Mistral AI | 1 214 | 1 204–1 223 | 3 942 | Apache 2.0 |
| 297 | GPT-4 · 0613 | OpenAI | 1 214 | 1 207–1 221 | 22 099 | Proprietary |
| 298 | Nova Lite 1.0 | Amazon | 1 211 | 1 203–1 219 | 5 254 | Proprietary |
| 299 | GLM-4 | Z.ai | 1 210 | 1 198–1 222 | 2 609 | Proprietary |
| 300 | Llama 3 70B Instruct | Meta | 1 206 | 1 200–1 212 | 37 706 | Llama 3 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.