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 «испанский», 283 строк. Оценка и интервал — из источника; интервал не равен диапазону мест.
| Место | Модель | Разработчик | Оценка | Интервал | Голоса | Лицензия |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 в режиме max | Anthropic | 1 522 | 1 495–1 549 | 541 | Proprietary |
| 2 | Claude Opus 4.6 | Anthropic | 1 512 | 1 499–1 525 | 2 636 | Proprietary |
| 3 | Claude Opus 5 в режиме high | Anthropic | 1 511 | 1 492–1 529 | 1 156 | Proprietary |
| 4 | Claude Fable 5 | Anthropic | 1 509 | 1 489–1 528 | 981 | Proprietary |
| 5 | Claude Opus 4.6 в режиме high | Anthropic | 1 503 | 1 489–1 516 | 2 530 | Proprietary |
| 6 | Qwen3.8 Max (0902) | Alibaba Qwen | 1 501 | 1 472–1 529 | 464 | Proprietary |
| 7 | Claude Opus 4.7 в режиме high | Anthropic | 1 497 | 1 483–1 512 | 1 964 | Proprietary |
| 8 | Muse Spark 1.1 | Meta | 1 480 | 1 458–1 501 | 812 | Proprietary |
| 9 | Gemini 3.1 Pro Preview | 1 479 | 1 468–1 491 | 3 518 | Proprietary | |
| 10 | Kimi K3 в режиме max | Moonshot AI | 1 479 | 1 455–1 502 | 633 | Kimi K3 license |
| 11 | ERNIE 5.1 | Baidu | 1 477 | 1 459–1 496 | 1 257 | Proprietary |
| 12 | GLM 5.2 в режиме max | Z.ai | 1 477 | 1 458–1 496 | 1 080 | MIT |
| 13 | GLM 5.3 Flash | Z.ai | 1 477 | 1 442–1 512 | 300 | MIT |
| 14 | Gemini 3.6 Flash в режиме high | 1 477 | 1 452–1 501 | 658 | Proprietary | |
| 15 | ERNIE 5.0 · preview-1203 | Baidu | 1 476 | 1 439–1 513 | 235 | Proprietary |
| 16 | Gemini 3.5 Flash в режиме medium | 1 475 | 1 456–1 494 | 1 016 | Proprietary | |
| 17 | Gemini 2.5 Pro | 1 474 | 1 462–1 485 | 3 413 | Proprietary | |
| 18 | Gemini 3.5 Flash в режиме high | 1 473 | 1 455–1 491 | 1 154 | Proprietary | |
| 19 | Muse Spark | Meta | 1 472 | 1 446–1 499 | 500 | Proprietary |
| 20 | Claude Opus 4.7 | Anthropic | 1 472 | 1 458–1 486 | 2 031 | Proprietary |
| 21 | Gemini 3 Flash Preview | 1 470 | 1 451–1 489 | 987 | Proprietary | |
| 22 | Claude Opus 4.8 в режиме high | Anthropic | 1 470 | 1 454–1 486 | 1 562 | Proprietary |
| 23 | Qwen3.5 Max | Alibaba Qwen | 1 469 | 1 448–1 491 | 781 | Proprietary |
| 24 | Gemini 3 Pro | 1 468 | 1 451–1 486 | 1 207 | Proprietary | |
| 25 | GPT-5.5 в режиме high | OpenAI | 1 466 | 1 452–1 481 | 1 899 | Proprietary |
| 26 | MiMo-V2.5-Pro | Xiaomi | 1 466 | 1 451–1 481 | 1 863 | MIT |
| 27 | GLM 5.1 | Z.ai | 1 466 | 1 451–1 482 | 1 646 | MIT |
| 28 | Claude Sonnet 4.6 | Anthropic | 1 465 | 1 451–1 479 | 2 149 | Proprietary |
| 29 | Qwen3 Max · preview | Alibaba Qwen | 1 463 | 1 443–1 484 | 838 | Proprietary |
| 30 | Kimi K2.6 | Moonshot AI | 1 463 | 1 446–1 480 | 1 320 | Modified MIT |
| 31 | GPT-5.5 · 5.5 | OpenAI | 1 459 | 1 445–1 474 | 1 990 | Proprietary |
| 32 | Seed 2.0 Pro | ByteDance Seed | 1 458 | 1 445–1 472 | 2 492 | Proprietary |
| 33 | Qwen3.8 27B | Alibaba Qwen | 1 456 | 1 421–1 491 | 294 | Apache 2.0 |
| 34 | Qwen3.7 Plus | Alibaba Qwen | 1 456 | 1 438–1 474 | 1 221 | Proprietary |
| 35 | GPT-5.4 в режиме high | OpenAI | 1 456 | 1 441–1 470 | 2 020 | Proprietary |
| 36 | Nemotron 3 Ultra | NVIDIA | 1 456 | 1 426–1 485 | 432 | OpenMDW-1.1 |
| 37 | MiMo-V2 Pro | Xiaomi | 1 455 | 1 434–1 477 | 827 | Proprietary |
| 38 | GPT-5.6 Terra в режиме xhigh | OpenAI | 1 455 | 1 434–1 477 | 786 | Proprietary |
| 39 | Claude Sonnet 4.5 | Anthropic | 1 455 | 1 442–1 467 | 2 442 | Proprietary |
| 40 | DeepSeek V4 Pro 0423 | DeepSeek | 1 454 | 1 439–1 470 | 1 690 | MIT |
| 41 | Claude Opus 4.5 в режиме high | Anthropic | 1 454 | 1 436–1 472 | 1 120 | Proprietary |
| 42 | Gemma 4 31B | 1 453 | 1 417–1 490 | 245 | Apache 2.0 | |
| 43 | Claude Opus 4.5 | Anthropic | 1 453 | 1 440–1 467 | 2 112 | Proprietary |
| 44 | GLM 4.5 | Z.ai | 1 453 | 1 430–1 477 | 674 | MIT |
| 45 | Qwen3.6 Max Preview | Alibaba Qwen | 1 452 | 1 415–1 489 | 250 | Proprietary |
| 46 | GLM 5 | Z.ai | 1 452 | 1 434–1 470 | 1 118 | MIT |
| 47 | Claude Opus 4.8 | Anthropic | 1 452 | 1 436–1 468 | 1 604 | Proprietary |
| 48 | GPT-5.4 | OpenAI | 1 451 | 1 437–1 466 | 1 985 | Proprietary |
| 49 | Kimi K2.5 · thinking | Moonshot AI | 1 451 | 1 437–1 464 | 2 468 | Modified MIT |
| 50 | DeepSeek V4 Pro 0423 в режиме high | DeepSeek | 1 450 | 1 434–1 466 | 1 655 | MIT |
| 51 | GPT-5.6 Luna в режиме xhigh | OpenAI | 1 449 | 1 427–1 471 | 793 | Proprietary |
| 52 | Grok 4.5 | xAI | 1 449 | 1 427–1 470 | 826 | Proprietary |
| 53 | Claude Opus 4.1 · 20250805 | Anthropic | 1 447 | 1 434–1 460 | 2 277 | Proprietary |
| 54 | ERNIE 5.0 · 0110 | Baidu | 1 446 | 1 430–1 463 | 1 344 | Proprietary |
| 55 | GPT-5.6 Sol в режиме xhigh | OpenAI | 1 446 | 1 424–1 468 | 771 | Proprietary |
| 56 | Grok 4.20 Multi-Agent | xAI | 1 445 | 1 431–1 460 | 1 963 | Proprietary |
| 57 | Amazon Nova Experimental Chat · 11-10 | Amazon | 1 445 | 1 424–1 467 | 787 | Proprietary |
| 58 | LongCat-Flash-Chat · chat | Meituan | 1 445 | 1 415–1 476 | 401 | MIT |
| 59 | Qwen3.5 397B A17B | Alibaba Qwen | 1 445 | 1 432–1 458 | 2 448 | Apache 2.0 |
| 60 | Inkling | Thinking Machines Lab | 1 444 | 1 422–1 466 | 763 | Apache 2.0 |
| 61 | Grok 4.20 · beta-0309-reasoning | xAI | 1 444 | 1 430–1 459 | 1 996 | Proprietary |
| 62 | GLM 5.3 в режиме max | Z.ai | 1 444 | 1 412–1 477 | 295 | MIT |
| 63 | Claude Opus 4.1 · 20250805-thinking-16k | Anthropic | 1 444 | 1 427–1 461 | 1 287 | Proprietary |
| 64 | Kimi K2.5 · instant | Moonshot AI | 1 444 | 1 412–1 475 | 351 | Modified MIT |
| 65 | GLM 5V Turbo | Z.ai | 1 444 | 1 407–1 480 | 290 | Proprietary |
| 66 | Hy3 | Tencent | 1 443 | 1 409–1 476 | 293 | Apache 2.0 |
| 67 | Grok 4.20 · beta1 | xAI | 1 441 | 1 421–1 461 | 917 | Proprietary |
| 68 | DeepSeek V3.2 Exp · exp | DeepSeek | 1 440 | 1 409–1 471 | 349 | MIT |
| 69 | Claude Sonnet 4.5 в режиме high | Anthropic | 1 439 | 1 427–1 452 | 2 530 | Proprietary |
| 70 | Claude Sonnet 5 в режиме high | Anthropic | 1 439 | 1 420–1 459 | 1 009 | Proprietary |
| 71 | Mistral Large 3 2512 | Mistral AI | 1 439 | 1 426–1 453 | 2 108 | Apache 2.0 |
| 72 | Grok 4.1 · 4.1-thinking | xAI | 1 439 | 1 424–1 453 | 1 903 | Proprietary |
| 73 | DeepSeek V4 Flash 0423 | DeepSeek | 1 438 | 1 422–1 454 | 1 546 | MIT |
| 74 | ChatGPT-4o (latest) | OpenAI | 1 436 | 1 422–1 450 | 1 888 | Proprietary |
| 75 | Qwen3 Next 80B A3B Instruct | Alibaba Qwen | 1 435 | 1 412–1 458 | 684 | Apache 2.0 |
| 76 | GLM 4.6 | Z.ai | 1 435 | 1 415–1 454 | 1 004 | MIT |
| 77 | GPT-5.2 Chat | OpenAI | 1 434 | 1 416–1 452 | 1 160 | Proprietary |
| 78 | Mistral Medium 3.1 | Mistral AI | 1 434 | 1 422–1 446 | 2 827 | Proprietary |
| 79 | Qwen3.6 Plus | Alibaba Qwen | 1 433 | 1 417–1 449 | 1 557 | Proprietary |
| 80 | Gemini 3.5 Flash Lite | 1 433 | 1 409–1 457 | 666 | Proprietary | |
| 81 | Inkling Small | Thinking Machines Lab | 1 432 | 1 405–1 459 | 493 | Apache 2.0 |
| 82 | GPT-5.1 в режиме high | OpenAI | 1 432 | 1 414–1 449 | 1 190 | Proprietary |
| 83 | Gemma 4 26B A4B | 1 431 | 1 390–1 473 | 201 | Apache 2.0 | |
| 84 | Gemini 3 Flash Preview в режиме minimal | 1 431 | 1 418–1 443 | 2 702 | Proprietary | |
| 85 | MiMo-V2 Omni | Xiaomi | 1 430 | 1 406–1 454 | 629 | Proprietary |
| 86 | GLM 4.7 | Z.ai | 1 430 | 1 399–1 461 | 342 | MIT |
| 87 | GPT-5.5 · 5.5-instant | OpenAI | 1 429 | 1 407–1 452 | 783 | Proprietary |
| 88 | DeepSeek V3.1 · v3.1 | DeepSeek | 1 429 | 1 399–1 459 | 409 | MIT |
| 89 | Qwen3 VL 235B A22B Instruct | Alibaba Qwen | 1 428 | 1 400–1 457 | 454 | Apache 2.0 |
| 90 | MiniMax M3 | MiniMax | 1 428 | 1 412–1 445 | 1 488 | MiniMax Community License |
| 91 | Qwen3 235B A22B Instruct 2507 | Alibaba Qwen | 1 427 | 1 415–1 439 | 2 884 | Apache 2.0 |
| 92 | DeepSeek V3.2 · v3.2 | DeepSeek | 1 425 | 1 409–1 441 | 1 499 | MIT |
| 93 | Grok 4.1 · 4.1 | xAI | 1 425 | 1 411–1 439 | 1 940 | Proprietary |
| 94 | Gemini 3.1 Flash Lite Preview | 1 425 | 1 410–1 439 | 1 960 | Proprietary | |
| 95 | DeepSeek V4 Pro 0813 в режиме high | DeepSeek | 1 422 | 1 388–1 456 | 269 | MIT |
| 96 | DeepSeek V4 Flash 0423 в режиме high | DeepSeek | 1 421 | 1 404–1 438 | 1 438 | MIT |
| 97 | ERNIE 5.0 · preview-1022 | Baidu | 1 421 | 1 376–1 466 | 166 | Proprietary |
| 98 | MiMo-V2 Flash · flash (non-thinking) | Xiaomi | 1 421 | 1 405–1 437 | 1 502 | MIT |
| 99 | Gemini 2.5 Flash · flash | 1 420 | 1 409–1 432 | 3 384 | Proprietary | |
| 100 | Qwen3 Max · 2025-09-23 | Alibaba Qwen | 1 420 | 1 392–1 449 | 435 | Proprietary |
| 101 | GPT-5.1 | OpenAI | 1 420 | 1 403–1 437 | 1 284 | Proprietary |
| 102 | Step 3.5 Flash | StepFun | 1 420 | 1 406–1 434 | 1 989 | Apache 2.0 |
| 103 | Qwen3.5-122B-A10B | Alibaba Qwen | 1 419 | 1 400–1 438 | 930 | Apache 2.0 |
| 104 | Grok 4.6 в режиме high | xAI | 1 419 | 1 388–1 450 | 369 | Proprietary |
| 105 | Grok 4 | xAI | 1 418 | 1 398–1 437 | 988 | Proprietary |
| 106 | Claude Haiku 4.5 | Anthropic | 1 417 | 1 406–1 429 | 3 789 | Proprietary |
| 107 | DeepSeek V3.2 · v3.2-thinking | DeepSeek | 1 417 | 1 401–1 433 | 1 351 | MIT |
| 108 | Grok 3 | xAI | 1 417 | 1 386–1 447 | 376 | Proprietary |
| 109 | GPT-5.2 в режиме high | OpenAI | 1 416 | 1 401–1 432 | 1 517 | Proprietary |
| 110 | Nemotron 3 Super | NVIDIA | 1 415 | 1 384–1 447 | 347 | NVIDIA Open Model |
| 111 | Grok 4 Fast · chat | xAI | 1 415 | 1 377–1 454 | 255 | Proprietary |
| 112 | Grok 4.1 Fast | xAI | 1 414 | 1 398–1 429 | 1 612 | Proprietary |
| 113 | MiMo-V2.5 | Xiaomi | 1 414 | 1 397–1 431 | 1 414 | MIT |
| 114 | DeepSeek V3.1 · v3.1-thinking | DeepSeek | 1 412 | 1 382–1 442 | 389 | MIT |
| 115 | DeepSeek V3.2 Exp · exp-thinking | DeepSeek | 1 410 | 1 377–1 443 | 310 | MIT |
| 116 | LongCat-Flash-Chat · chat-2602-exp | Meituan | 1 409 | 1 389–1 429 | 887 | Proprietary |
| 117 | Qwen3.5-27B | Alibaba Qwen | 1 409 | 1 389–1 428 | 924 | Apache 2.0 |
| 118 | R1 0528 | DeepSeek | 1 409 | 1 373–1 444 | 260 | MIT |
| 119 | Kimi K2 Thinking | Moonshot AI | 1 408 | 1 394–1 423 | 1 958 | Modified MIT |
| 120 | MiniMax M2.7 | MiniMax | 1 405 | 1 391–1 419 | 2 248 | Modified MIT |
| 121 | GPT-5.4 Mini в режиме high | OpenAI | 1 404 | 1 389–1 419 | 1 856 | Proprietary |
| 122 | Gemini 2.5 Flash · flash-preview-09-2025 | 1 404 | 1 384–1 424 | 971 | Proprietary | |
| 123 | Qwen3 VL 235B A22B Thinking | Alibaba Qwen | 1 404 | 1 372–1 435 | 375 | Apache 2.0 |
| 124 | Grok 4 Fast · reasoning | xAI | 1 403 | 1 379–1 428 | 591 | Proprietary |
| 125 | Grok 4.3 | xAI | 1 403 | 1 389–1 418 | 1 968 | Proprietary |
| 126 | Qwen3 30B A3B Instruct 2507 | Alibaba Qwen | 1 403 | 1 381–1 426 | 712 | Apache 2.0 |
| 127 | Ling-flash-2.0 | inclusionAI | 1 403 | 1 367–1 439 | 276 | MIT |
| 128 | Qwen3 235B A22B · a22b-no-thinking | Alibaba Qwen | 1 403 | 1 381–1 424 | 762 | Apache 2.0 |
| 129 | Qwen3.5-Flash | Alibaba Qwen | 1 401 | 1 386–1 415 | 1 995 | Proprietary |
| 130 | Mistral Medium 3.5 | Mistral AI | 1 401 | 1 369–1 432 | 377 | Modified MIT |
| 131 | GPT-5 | OpenAI | 1 399 | 1 379–1 420 | 898 | Proprietary |
| 132 | Gemini 2.5 Flash Lite · 09-2025-no-thinking | 1 398 | 1 382–1 414 | 1 413 | Proprietary | |
| 133 | GPT-5.2 | OpenAI | 1 397 | 1 384–1 411 | 2 503 | Proprietary |
| 134 | MiniMax M2.1 | MiniMax | 1 396 | 1 369–1 424 | 477 | MIT |
| 135 | Hy3 preview | Tencent | 1 395 | 1 355–1 435 | 207 | tencent-hunyuan-community |
| 136 | o3 | OpenAI | 1 395 | 1 377–1 413 | 1 212 | Proprietary |
| 137 | MiMo-V2 Flash · flash (thinking) | Xiaomi | 1 394 | 1 365–1 423 | 379 | MIT |
| 138 | Qwen3 235B A22B Thinking 2507 | Alibaba Qwen | 1 394 | 1 353–1 435 | 184 | Apache 2.0 |
| 139 | Qwen3.5-35B-A3B | Alibaba Qwen | 1 390 | 1 370–1 409 | 988 | Apache 2.0 |
| 140 | Amazon Nova Experimental Chat · 26-01-10 | Amazon | 1 389 | 1 349–1 429 | 200 | Proprietary |
| 141 | Claude Opus 4 · 20250514-thinking-16k | Anthropic | 1 389 | 1 368–1 409 | 854 | Proprietary |
| 142 | GPT-5 в режиме high | OpenAI | 1 387 | 1 367–1 407 | 953 | Proprietary |
| 143 | gpt-oss-120b | OpenAI | 1 387 | 1 366–1 408 | 853 | Apache 2.0 |
| 144 | GLM 4.5 Air | Z.ai | 1 386 | 1 364–1 407 | 822 | MIT |
| 145 | Kimi K2 0905 | Moonshot AI | 1 382 | 1 351–1 414 | 388 | Modified MIT |
| 146 | Ring-flash-2.0 | inclusionAI | 1 381 | 1 347–1 416 | 306 | MIT |
| 147 | Grok 3 Mini | xAI | 1 379 | 1 350–1 408 | 392 | Proprietary |
| 148 | R1 | DeepSeek | 1 378 | 1 325–1 431 | 114 | MIT |
| 149 | Amazon Nova Experimental Chat · 10-20 | Amazon | 1 377 | 1 343–1 411 | 312 | Proprietary |
| 150 | Nemotron 3 Nano 30B A3B | NVIDIA | 1 377 | 1 345–1 409 | 352 | NVIDIA Open Model |
| 151 | Qwen2.5 Max | Alibaba Qwen | 1 377 | 1 341–1 412 | 268 | Proprietary |
| 152 | GPT-4.1 | OpenAI | 1 374 | 1 356–1 392 | 1 127 | Proprietary |
| 153 | Step 3 | StepFun | 1 372 | 1 330–1 414 | 209 | Apache 2.0 |
| 154 | Nova 2 Lite | Amazon | 1 371 | 1 332–1 411 | 231 | Proprietary |
| 155 | GPT-5.4 Nano в режиме high | OpenAI | 1 370 | 1 354–1 385 | 1 877 | Proprietary |
| 156 | GPT-5.3 Chat | OpenAI | 1 369 | 1 352–1 387 | 1 262 | Proprietary |
| 157 | Qwen3 235B A22B · a22b | Alibaba Qwen | 1 366 | 1 339–1 393 | 468 | Apache 2.0 |
| 158 | Gemini 2.0 Flash | 1 363 | 1 338–1 388 | 553 | Proprietary | |
| 159 | Mistral Medium 3 | Mistral AI | 1 363 | 1 338–1 388 | 564 | Proprietary |
| 160 | Gemini 2.5 Flash Lite · 06-17-thinking | 1 363 | 1 341–1 384 | 799 | Proprietary | |
| 161 | GPT-5 Nano в режиме high | OpenAI | 1 361 | 1 322–1 400 | 236 | Proprietary |
| 162 | Qwen3 Coder 480B A35B | Alibaba Qwen | 1 360 | 1 336–1 383 | 685 | Apache 2.0 |
| 163 | Grok 3 Mini в режиме high | xAI | 1 358 | 1 326–1 390 | 323 | Proprietary |
| 164 | DeepSeek V3 0324 | DeepSeek | 1 358 | 1 339–1 377 | 983 | MIT |
| 165 | Claude Sonnet 4 · 20250514-thinking-32k | Anthropic | 1 357 | 1 335–1 378 | 831 | Proprietary |
| 166 | Claude Opus 4 · 20250514 | Anthropic | 1 356 | 1 337–1 376 | 998 | Proprietary |
| 167 | GPT-5 Mini в режиме high | OpenAI | 1 356 | 1 335–1 377 | 852 | Proprietary |
| 168 | Qwen3 Next 80B A3B Thinking | Alibaba Qwen | 1 355 | 1 327–1 384 | 455 | Apache 2.0 |
| 169 | GLM 4.7 Flash | Z.ai | 1 355 | 1 327–1 384 | 438 | MIT |
| 170 | Trinity Large | Arcee AI | 1 354 | 1 335–1 374 | 1 029 | Apache 2.0 |
| 171 | MiniMax M2.5 | MiniMax | 1 354 | 1 337–1 371 | 1 461 | Modified MIT |
| 172 | MiniMax M1 | MiniMax | 1 353 | 1 332–1 375 | 823 | Apache 2.0 |
| 173 | QwQ 32B | Alibaba Qwen | 1 352 | 1 323–1 381 | 397 | Apache 2.0 |
| 174 | DeepSeek V3 | DeepSeek | 1 352 | 1 304–1 400 | 127 | DeepSeek |
| 175 | Gemma 3 27B | 1 349 | 1 329–1 369 | 883 | Gemma | |
| 176 | Claude Sonnet 4 · 20250514 | Anthropic | 1 348 | 1 327–1 369 | 888 | Proprietary |
| 177 | o4 Mini | OpenAI | 1 346 | 1 327–1 366 | 996 | Proprietary |
| 178 | Command A | Cohere | 1 345 | 1 328–1 363 | 1 175 | CC-BY-NC-4.0 |
| 179 | Mistral Small 3.2 24B | Mistral AI | 1 345 | 1 315–1 374 | 420 | Apache 2.0 |
| 180 | o1 · 2024-12-17 | OpenAI | 1 345 | 1 296–1 393 | 142 | Proprietary |
| 181 | OLMo 3.1 32B Instruct | Ai2 | 1 336 | 1 305–1 367 | 368 | Apache 2.0 |
| 182 | GLM 4.5V | Z.ai | 1 335 | 1 290–1 380 | 186 | MIT |
| 183 | Trinity Large Thinking | Arcee AI | 1 333 | 1 312–1 354 | 910 | Apache 2.0 |
| 184 | Nemotron 3.5 Lightning | NVIDIA | 1 332 | 1 293–1 371 | 238 | OpenMDW-1.1 |
| 185 | Kimi K2 0711 | Moonshot AI | 1 329 | 1 304–1 354 | 618 | Modified MIT |
| 186 | MiniMax M2 | MiniMax | 1 325 | 1 282–1 368 | 204 | Apache 2.0 |
| 187 | o3 Mini High | OpenAI | 1 321 | 1 270–1 371 | 126 | Proprietary |
| 188 | GPT-4.1 Mini | OpenAI | 1 319 | 1 298–1 340 | 804 | Proprietary |
| 189 | Qwen3 30B A3B | Alibaba Qwen | 1 317 | 1 289–1 344 | 479 | Apache 2.0 |
| 190 | Yi-Lightning | 01.AI | 1 315 | 1 283–1 347 | 341 | Proprietary |
| 191 | Gemini 2.0 Flash-Lite | 1 313 | 1 269–1 357 | 157 | Proprietary | |
| 192 | GLM-4-Plus | Z.ai | 1 310 | 1 278–1 342 | 332 | Proprietary |
| 193 | Gemini 1.5 Pro · 1.5-pro-002 | 1 310 | 1 281–1 339 | 405 | Proprietary | |
| 194 | o3 Mini | OpenAI | 1 309 | 1 288–1 329 | 916 | Proprietary |
| 195 | Gemma 3n E4B | 1 305 | 1 276–1 333 | 451 | Gemma | |
| 196 | o1-mini | OpenAI | 1 302 | 1 274–1 330 | 425 | Proprietary |
| 197 | o1 · preview | OpenAI | 1 301 | 1 270–1 333 | 347 | Proprietary |
| 198 | Claude 3.7 Sonnet · 20250219-thinking-32k | Anthropic | 1 298 | 1 273–1 323 | 611 | Proprietary |
| 199 | Llama 4 Maverick | Meta | 1 291 | 1 270–1 313 | 818 | Llama 4 |
| 200 | GPT-4o (2024-05-13) | OpenAI | 1 291 | 1 271–1 312 | 1 712 | Proprietary |
| 201 | Claude 3.7 Sonnet · 20250219 | Anthropic | 1 289 | 1 265–1 314 | 623 | Proprietary |
| 202 | Qwen Max | Alibaba Qwen | 1 289 | 1 251–1 328 | 222 | Qwen |
| 203 | Claude 3.5 Sonnet · 20240620 | Anthropic | 1 289 | 1 266–1 311 | 1 123 | Proprietary |
| 204 | Claude 3.5 Sonnet · 20241022 | Anthropic | 1 285 | 1 265–1 304 | 968 | Proprietary |
| 205 | OLMo 3.1 32B Think | Ai2 | 1 284 | 1 240–1 328 | 191 | Apache 2.0 |
| 206 | Mistral Small 3.1 24B | Mistral AI | 1 283 | 1 261–1 305 | 803 | Apache 2.0 |
| 207 | Grok 2 | xAI | 1 280 | 1 256–1 305 | 671 | Proprietary |
| 208 | Grok 2 Mini | xAI | 1 278 | 1 253–1 303 | 575 | Proprietary |
| 209 | Llama 4 Scout | Meta | 1 276 | 1 252–1 300 | 710 | Llama |
| 210 | GPT-4o-mini (2024-07-18) | OpenAI | 1 275 | 1 252–1 299 | 741 | Proprietary |
| 211 | Athene V2 Chat | Nexusflow | 1 274 | 1 234–1 314 | 180 | NexusFlow |
| 212 | Magistral Medium | Mistral AI | 1 269 | 1 230–1 309 | 268 | Proprietary |
| 213 | Llama 3.3 70B Instruct | Meta | 1 269 | 1 245–1 293 | 635 | Llama-3.3 |
| 214 | GPT-4o (2024-08-06) | OpenAI | 1 268 | 1 241–1 295 | 546 | Proprietary |
| 215 | gpt-oss-20b | OpenAI | 1 267 | 1 229–1 306 | 284 | Apache 2.0 |
| 216 | Mistral Large · 2411 | Mistral AI | 1 267 | 1 222–1 313 | 153 | MRL |
| 217 | Claude 3.5 Haiku | Anthropic | 1 260 | 1 241–1 280 | 962 | Proprietary |
| 218 | GPT-4 Turbo Preview · 1106-preview | OpenAI | 1 260 | 1 238–1 282 | 1 472 | Proprietary |
| 219 | GPT-4 Turbo | OpenAI | 1 259 | 1 238–1 281 | 1 430 | Proprietary |
| 220 | Llama 3.1 405B Instruct · bf16 | Meta | 1 259 | 1 229–1 290 | 348 | Llama 3.1 Community |
| 221 | Athene 70B | Nexusflow | 1 259 | 1 223–1 294 | 271 | CC-BY-NC-4.0 |
| 222 | Granite 4.0 H Small | IBM Granite | 1 258 | 1 212–1 304 | 192 | Apache 2.0 |
| 223 | Qwen2.5 72B Instruct | Alibaba Qwen | 1 255 | 1 224–1 286 | 347 | Qwen |
| 224 | Llama 3.1 70B Instruct | Meta | 1 252 | 1 227–1 278 | 606 | Llama 3.1 Community |
| 225 | Llama 3.1 405B Instruct · fp8 | Meta | 1 252 | 1 227–1 278 | 652 | Llama 3.1 Community |
| 226 | DeepSeek V2.5 | DeepSeek | 1 247 | 1 213–1 282 | 293 | DeepSeek |
| 227 | GPT-4 Turbo Preview · 0125-preview | OpenAI | 1 247 | 1 224–1 269 | 1 219 | Proprietary |
| 228 | Claude 3 Opus | Anthropic | 1 246 | 1 227–1 265 | 2 569 | Proprietary |
| 229 | Gemini 1.5 Pro · 1.5-pro-001 | 1 242 | 1 220–1 264 | 1 300 | Proprietary | |
| 230 | Gemini 1.5 Flash · 002 | 1 242 | 1 208–1 276 | 283 | Proprietary | |
| 231 | Gemini 1.5 Pro · advanced-0514 | 1 241 | 1 217–1 266 | 898 | Proprietary | |
| 232 | Llama 3 70B Instruct | Meta | 1 240 | 1 221–1 260 | 2 705 | Llama 3 Community |
| 233 | Mistral Large 2407 | Mistral AI | 1 236 | 1 210–1 262 | 584 | Mistral Research |
| 234 | Phi 4 | Microsoft | 1 234 | 1 182–1 285 | 127 | MIT |
| 235 | Gemma 2 27B | 1 228 | 1 205–1 250 | 845 | Gemma license | |
| 236 | Nova Lite 1.0 | Amazon | 1 225 | 1 172–1 278 | 107 | Proprietary |
| 237 | Nova Micro 1.0 | Amazon | 1 224 | 1 179–1 270 | 144 | Proprietary |
| 238 | Gemini 1.5 Flash · 001 | 1 224 | 1 201–1 246 | 1 077 | Proprietary | |
| 239 | Gemini 1.5 Flash-8B | 1 212 | 1 180–1 243 | 323 | Proprietary | |
| 240 | Claude 3 Sonnet | Anthropic | 1 204 | 1 182–1 226 | 1 448 | Proprietary |
| 241 | Gemma 2 9B | 1 200 | 1 174–1 225 | 614 | Gemma license | |
| 242 | GPT-4 · 0314 | OpenAI | 1 196 | 1 168–1 224 | 647 | Proprietary |
| 243 | Nemotron-4 340B Instruct | NVIDIA | 1 194 | 1 163–1 226 | 367 | NVIDIA Open Model |
| 244 | Mistral Large · 2402 | Mistral AI | 1 194 | 1 169–1 219 | 886 | Proprietary |
| 245 | Aya Expanse 32B | Cohere | 1 192 | 1 156–1 228 | 256 | CC-BY-NC-4.0 |
| 246 | Command R+ | Cohere | 1 188 | 1 166–1 211 | 1 174 | CC-BY-NC-4.0 |
| 247 | Nova Pro 1.0 | Amazon | 1 182 | 1 132–1 231 | 150 | Proprietary |
| 248 | Llama 3 8B Instruct | Meta | 1 172 | 1 151–1 193 | 1 744 | Llama 3 Community |
| 249 | Llama 3.1 8B Instruct | Meta | 1 169 | 1 141–1 196 | 565 | Llama 3.1 Community |
| 250 | Qwen2 72B Instruct | Alibaba Qwen | 1 168 | 1 140–1 196 | 596 | Qianwen LICENSE |
| 251 | GPT-4 · 0613 | OpenAI | 1 167 | 1 145–1 190 | 1 364 | Proprietary |
| 252 | Claude 3 Haiku | Anthropic | 1 165 | 1 144–1 186 | 1 704 | Proprietary |
| 253 | DeepSeek Coder V2 | DeepSeek | 1 152 | 1 115–1 190 | 245 | DeepSeek License |
| 254 | Command R | Cohere | 1 150 | 1 123–1 177 | 698 | CC-BY-NC-4.0 |
| 255 | Mixtral 8x22B Instruct | Mistral AI | 1 150 | 1 124–1 176 | 826 | Apache 2.0 |
| 256 | Reka Flash 21B · 20240226-online | Reka AI | 1 148 | 1 107–1 190 | 217 | Proprietary |
| 257 | Mistral Medium (2023) | Mistral AI | 1 143 | 1 111–1 175 | 439 | Proprietary |
| 258 | Llama 2 70B Chat | Meta | 1 142 | 1 113–1 171 | 506 | Llama 2 Community |
| 259 | Qwen1.5 110B Chat | Alibaba Qwen | 1 141 | 1 111–1 172 | 487 | Qianwen LICENSE |
| 260 | Gemma 2 2B | 1 139 | 1 112–1 167 | 544 | Gemma license | |
| 261 | Reka Flash 21B · 20240226 | Reka AI | 1 131 | 1 099–1 163 | 409 | Proprietary |
| 262 | Yi-1.5 34B Chat | 01.AI | 1 120 | 1 089–1 151 | 458 | Apache-2.0 |
| 263 | GPT-3.5 Turbo · 0125 | OpenAI | 1 120 | 1 095–1 145 | 895 | Proprietary |
| 264 | Gemini 1.0 Pro | 1 118 | 1 077–1 160 | 213 | Proprietary | |
| 265 | Phi-3 Small 8K Instruct | Microsoft | 1 110 | 1 073–1 148 | 272 | MIT |
| 266 | Mixtral 8x7B Instruct | Mistral AI | 1 110 | 1 086–1 134 | 1 065 | Apache 2.0 |
| 267 | Qwen1.5 72B Chat | Alibaba Qwen | 1 110 | 1 080–1 139 | 501 | Qianwen LICENSE |
| 268 | Phi-3 Medium 4K Instruct | Microsoft | 1 094 | 1 063–1 125 | 419 | MIT |
| 269 | GPT-3.5 Turbo · 1106 | OpenAI | 1 092 | 1 051–1 133 | 260 | Proprietary |
| 270 | Qwen1.5 32B Chat | Alibaba Qwen | 1 088 | 1 056–1 120 | 369 | Qianwen LICENSE |
| 271 | Snowflake Arctic Instruct | Snowflake | 1 087 | 1 057–1 118 | 500 | Apache 2.0 |
| 272 | Llama 2 13B Chat | Meta | 1 086 | 1 046–1 125 | 262 | Llama 2 Community |
| 273 | Qwen1.5 14B Chat | Alibaba Qwen | 1 085 | 1 049–1 120 | 293 | Qianwen LICENSE |
| 274 | Phi-3 Mini 4K Instruct | Microsoft | 1 084 | 1 051–1 117 | 410 | MIT |
| 275 | Vicuna 33B | LMSYS | 1 079 | 1 041–1 117 | 275 | Non-commercial |
| 276 | Vicuna 13B | LMSYS | 1 076 | 1 027–1 125 | 161 | Llama 2 Community |
| 277 | Zephyr 7B Beta | Hugging Face H4 | 1 073 | 1 017–1 128 | 127 | MIT |
| 278 | Yi 34B Chat | 01.AI | 1 069 | 1 028–1 111 | 218 | Yi License |
| 279 | DBRX Instruct | Databricks | 1 063 | 1 030–1 096 | 404 | DBRX LICENSE |
| 280 | Phi-3 Mini 128K Instruct | Microsoft | 1 058 | 1 023–1 092 | 366 | MIT |
| 281 | Gemma 1.1 7B | 1 049 | 1 015–1 082 | 411 | Gemma license | |
| 282 | Mistral 7B Instruct | Mistral AI | 1 026 | 980–1 071 | 202 | Apache-2.0 |
| 283 | Llama 2 7B Chat | Meta | 1 007 | 956–1 057 | 143 | Llama 2 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.