Serverless GPU inference: 142 модели, $/1M tokens и pay-per-image
- Моделей
- 183
- Средняя цена
- ₽84/M
- Регион
- Зарубежный · сайт на английский
- Оплата
- USD (доллары), Pay-as-you-go
- Биллинг
- за токены
- API
- OpenAI-compatible
input_seconds (STT) и input_characters (TTS) — другие unit, не $/1M tokens
Обновлено
Профиль расчёта: 25M · 80/20 · ₽
Токенов/мес
In/Out
Валюта
Токенов/мес
In/Out
Валюта
25M токенов/мес · ввод 80% / вывод 20%Цены в ₽ · 1 USD = 85.6007 ₽ · 1 CNY = 12.7335 ₽ · 29.08.2026
183 моделей
Текст и чат
оплата за токены · 125
▼
Текст и чат
оплата за токены · 125
| Модель | Maker | Input | Output | TCO/мес |
|---|---|---|---|---|
| Embeddinggemma 300m | ₽0.17/M | ₽0.17/M | ₽3 | |
| all-MiniLM-L12-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| all-MiniLM-L6-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| all-mpnet-base-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| bge-base-en-v1.5 | baai | ₽0.43/M | ₽0.43/M | ₽9 |
| clip-ViT-B-32 | sentence transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| clip-ViT-B-32-multilingual-v1 | sentence transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| E5-Base-v2 | Intfloat | ₽0.43/M | ₽0.43/M | ₽9 |
| GTE-Base | Thenlper | ₽0.43/M | ₽0.43/M | ₽9 |
| multi-qa-mpnet-base-dot-v1 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| paraphrase-MiniLM-L6-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| text2vec-base-chinese | shibing624 | ₽0.43/M | ₽0.43/M | ₽9 |
| Bge En Icl | BAAI | ₽0.86/M | ₽0.86/M | ₽17 |
| Bge M3 | baai | ₽0.86/M | ₽0.86/M | ₽17 |
| Bge M3 Multi | BAAI | ₽0.86/M | ₽0.86/M | ₽17 |
| bge-large-en-v1.5 | baai | ₽0.86/M | ₽0.86/M | ₽17 |
| E5-Large-v2 | Intfloat | ₽0.86/M | ₽0.86/M | ₽17 |
| GTE-Large | Thenlper | ₽0.86/M | ₽0.86/M | ₽17 |
| Llama Nemotron Embed Vl 1b V2 | nvidia | ₽0.86/M | ₽0.86/M | ₽17 |
| Multilingual-E5-Large | Intfloat | ₽0.86/M | ₽0.86/M | ₽17 |
| Qwen3 Embedding 8B | qwen3 embedding 8b | ₽0.86/M | ₽0.86/M | ₽17 |
| qwen3-embedding-0.6b | qwen | ₽0.86/M | ₽0.86/M | ₽17 |
| Qwen3 Embedding 4B | qwen3 embedding 4b | ₽1.71/M | ₽1.71/M | ₽34 |
| Mistral-Nemo-Instruct-2407 | mistralai | ₽1.63/M | ₽2.57/M | ₽45 |
| Meta-Llama-3.1-8B-Instruct-Turbo | meta llama | ₽1.71/M | ₽3.42/M | ₽51 |
| gemma-4-E4B-it | ₽1.71/M | ₽8.56/M | ₽77 | |
| L3-8B-Lunaris-v1-Turbo | Sao10K | ₽3.42/M | ₽4.28/M | ₽90 |
| granite-4.2-3b | ibm granite | ₽2.57/M | ₽10.27/M | ₽103 |
| gpt oss 20b | openai | ₽2.57/M | ₽11.98/M | ₽111 |
| mistralai · mistral-small-24b-instruct-2501 | mistralai | ₽4.28/M | ₽6.85/M | ₽120 |
| Gemma 3 4B It | gemma 3 4b it | ₽4.28/M | ₽8.56/M | ₽128 |
| gpt oss 120b | openai | ₽3.17/M | ₽14.55/M | ₽136 |
| gemma 3 12b it | ₽4.28/M | ₽12.84/M | ₽150 | |
| Nvidia Nemotron 3 Nano 30B A3B | nvidia nemotron 3 nano 30b a3b | ₽4.28/M | ₽17.12/M | ₽171 |
| Ling-3.0-flash | inclusionai | ₽5.14/M | ₽15.41/M | ₽180 |
| Phi 4 | Microsoft | ₽5.99/M | ₽11.98/M | ₽180 |
| gemma 3 27b it | ₽6.85/M | ₽13.70/M | ₽205 | |
| granite-4.2-8b | ibm granite | ₽5.14/M | ₽21.40/M | ₽210 |
| deepseek v4 flash | deepseek | ₽6.85/M | ₽15.41/M | ₽214 |
| Mistral-Small-3.2-24B-Instruct-2506 | mistralai | ₽6.42/M | ₽17.12/M | ₽214 |
| Nemotron 3.5 Lightning | NVIDIA | ₽6.85/M | ₽17.12/M | ₽223 |
| qwen3.5 9b | qwen | ₽8.56/M | ₽12.84/M | ₽235 |
| qwen3 32b | qwen | ₽6.85/M | ₽23.97/M | ₽257 |
| gemma 4 26b a4b it | ₽5.99/M | ₽29.10/M | ₽265 | |
| glm 4.7 flash | z ai | ₽5.14/M | ₽34.24/M | ₽274 |
| gemma-4-31B-it-turbo | ₽7.70/M | ₽29.10/M | ₽300 | |
| Llama 4 Scout 17b 16e Instruct | meta | ₽8.56/M | ₽25.68/M | ₽300 |
| Llama-3.3-70B-Instruct-Turbo | meta llama | ₽8.56/M | ₽27.39/M | ₽308 |
| qwen3 14b | qwen | ₽10.27/M | ₽20.54/M | ₽308 |
| Nvidia Nemotron 3 Super 120B A12B | nvidia nemotron 3 super 120b a12b | ₽7.28/M | ₽34.24/M | ₽317 |
| Dola-Seed 2.0 Mini | bytedance | ₽8.56/M | ₽34.24/M | ₽342 |
| llama guard 4 12b | meta llama | ₽15.41/M | ₽15.41/M | ₽385 |
| gemma 4 31b it | ₽11.13/M | ₽32.53/M | ₽385 | |
| Qwen3 235B A22B Instruct | alibaba | ₽7.70/M | ₽47.08/M | ₽389 |
| qwen3 30b a3b | qwen | ₽10.27/M | ₽42.80/M | ₽419 |
| Nemotron-Content-Safety-3.5 | nvidia | ₽17.12/M | ₽17.12/M | ₽428 |
| glm 5.3 flash | z ai | ₽12.84/M | ₽42.80/M | ₽471 |
| Hy3 | hy3 | ₽11.98/M | ₽49.65/M | ₽488 |
| gpt-oss-120b-Turbo | openai | ₽12.84/M | ₽51.36/M | ₽514 |
| Qwen3 VL 30B | qwen3 vl 30b a3b | ₽12.84/M | ₽51.36/M | ₽514 |
| granite-4.2-30b | ibm granite | ₽13.70/M | ₽55.64/M | ₽552 |
| qwen3.6 35b a3b | qwen | ₽8.56/M | ₽81.32/M | ₽578 |
| deepseek v3.2 | deepseek | ₽22.26/M | ₽32.53/M | ₽608 |
| qwen3 next 80b a3b | qwen3 next 80b a3b | ₽7.70/M | ₽94.16/M | ₽625 |
| qwen3.5 35b a3b | qwen | ₽11.98/M | ₽85.60/M | ₽668 |
| Llama-4-Maverick-17B-128E-Instruct-FP8 | meta llama | ₽17.12/M | ₽68.48/M | ₽685 |
| qwen3 vl 235b a22b | qwen3 vl 235b a22b | ₽17.12/M | ₽75.33/M | ₽719 |
| gpt-oss-120b-Ultra | openai | ₽17.12/M | ₽81.32/M | ₽749 |
| gemma-4-31B-it-Ultra | ₽23.11/M | ₽65.06/M | ₽788 | |
| qwen 2.5 72b instruct | qwen | ₽30.82/M | ₽34.24/M | ₽788 |
| DeepSeek-V3 | deepseek ai | ₽20.54/M | ₽77.04/M | ₽796 |
| Step 3.7 Flash | step 3.7 flash | ₽17.12/M | ₽98.44/M | ₽835 |
| DeepSeek V3.1 | deepseek ai | ₽21.40/M | ₽81.32/M | ₽835 |
| Meta-Llama-3.1-70B-Instruct-Turbo | meta llama | ₽34.24/M | ₽34.24/M | ₽856 |
| minimax m2.7 | minimax | ₽21.40/M | ₽85.60/M | ₽856 |
| Mythomax L2 13B | Gryphe | ₽34.24/M | ₽34.24/M | ₽856 |
| Qwen3-Coder-480B-A35B-Instruct-Turbo | Qwen | ₽25.68/M | ₽85.60/M | ₽942 |
| minimax m3 | minimax | ₽23.97/M | ₽94.16/M | ₽950 |
| Muse Glimmer 30B | muse glimmer 30b | ₽25.68/M | ₽103/M | ₽1 027 |
| gemini 3.1 flash lite | ₽21.40/M | ₽128/M | ₽1 070 | |
| ByteDance Seed 1.8 | bytedance | ₽21.40/M | ₽171/M | ₽1 284 |
| Inkling Small | thinkingmachines | ₽38.52/M | ₽103/M | ₽1 284 |
| MiniMax-M2.7-Turbo | MiniMaxAI | ₽32.53/M | ₽146/M | ₽1 378 |
| glm 4.7 | z ai | ₽34.24/M | ₽150/M | ₽1 434 |
| Hermes 3 70B Instruct | nousresearch | ₽59.92/M | ₽59.92/M | ₽1 498 |
| qwen3.5 122b a10b | qwen | ₽24.82/M | ₽205/M | ₽1 524 |
| mimo v2.5 | mimo v2.5 | ₽34.24/M | ₽171/M | ₽1 541 |
| qwen3.5 27b | qwen | ₽22.26/M | ₽223/M | ₽1 558 |
| Gemini 2.5 Flash | gemini 2.5 flash | ₽25.68/M | ₽214/M | ₽1 584 |
| glm 4.6 | z ai | ₽42.80/M | ₽171/M | ₽1 712 |
| Kimi K2.5 | kimi k2.5 | ₽38.52/M | ₽193/M | ₽1 733 |
| DeepSeek R1 | deepseek r1 | ₽42.80/M | ₽184/M | ₽1 776 |
| NVIDIA-Nemotron-3-Ultra-550B-A55B | nvidia | ₽42.80/M | ₽188/M | ₽1 798 |
| L3.1-70B-Euryale-v2.2 | Sao10K | ₽72.76/M | ₽72.76/M | ₽1 819 |
| GLM-5 | glm 5 | ₽51.36/M | ₽178/M | ₽1 917 |
| qwen3.6 27b | qwen | ₽27.39/M | ₽274/M | ₽1 917 |
| qwen3.8 27b | qwen | ₽34.24/M | ₽257/M | ₽1 969 |
| qwen3.5 397b a17b | qwen | ₽38.52/M | ₽257/M | ₽2 054 |
| Dola-Seed 2.0 Pro | bytedance | ₽42.80/M | ₽257/M | ₽2 140 |
| seed-2.0-code | seed 2 0 code | ₽42.80/M | ₽257/M | ₽2 140 |
| Hermes 3 405B Instruct | nousresearch | ₽85.60/M | ₽85.60/M | ₽2 140 |
| glm 5.2 | z ai | ₽64.20/M | ₽205/M | ₽2 311 |
| kimi k2.7 code | moonshotai | ₽58.21/M | ₽291/M | ₽2 619 |
| kimi k2.6 | moonshotai | ₽64.20/M | ₽300/M | ₽2 782 |
| gemini 3.7 flash | ₽64.20/M | ₽321/M | ₽2 889 | |
| mimo v2.5 pro | mimo v2.5 pro | ₽85.60/M | ₽257/M | ₽2 996 |
| glm 5.1 | z ai | ₽89.88/M | ₽300/M | ₽3 296 |
| deepseek v4 pro | deepseek | ₽111/M | ₽223/M | ₽3 338 |
| Inkling | Thinking Machines | ₽81.32/M | ₽347/M | ₽3 360 |
| glm 5.3 | z ai | ₽103/M | ₽342/M | ₽3 766 |
| Claude Haiku 4.5 | claude haiku 4 5 | ₽85.60/M | ₽428/M | ₽3 852 |
| Qwen3 Max | qwen3 max | ₽103/M | ₽514/M | ₽4 622 |
| qwen3.8 max | qwen | ₽141/M | ₽424/M | ₽4 944 |
| Qwen3.8 2.4t A95B | qwen3.8 2.4t a95b | ₽171/M | ₽514/M | ₽5 992 |
| gemini 3.5 flash | ₽128/M | ₽770/M | ₽6 420 | |
| Gemini 2.5 Pro | gemini 2.5 pro | ₽107/M | ₽856/M | ₽6 420 |
| qwen3.7 max | qwen | ₽214/M | ₽642/M | ₽7 490 |
| claude sonnet 5 | anthropic | ₽171/M | ₽856/M | ₽7 704 |
| gemini 3.1 pro | ₽171/M | ₽1 027/M | ₽8 560 | |
| kimi k3 | moonshotai | ₽244/M | ₽1 220/M | ₽10 978 |
| Claude Sonnet 4.6 | claude sonnet 4 6 | ₽257/M | ₽1 284/M | ₽11 556 |
| Claude Opus 4.7 | claude opus 4 7 | ₽428/M | ₽2 140/M | ₽19 260 |
| claude opus 4.8 | anthropic | ₽428/M | ₽2 140/M | ₽19 260 |
| claude opus 5 | anthropic | ₽428/M | ₽2 140/M | ₽19 260 |
| claude fable 5 | anthropic | ₽856/M | ₽4 280/M | ₽38 520 |
Изображения
за изображение · 24
▼
Изображения
за изображение · 24
| Модель | Maker | Цена |
|---|---|---|
| Sdxl Turbo | stabilityai | ₽0.02/изобр. |
| Flux 1 Schnell | black forest labs | ₽0.04/изобр. |
| p-image | PrunaAI | ₽0.43/изобр. |
| FLUX.1 [dev] | black forest labs | ₽0.77/изобр. |
| Flux 2 Dev | black forest labs | ₽0.86/изобр. |
| FLUX.1-Kontext-dev | black forest labs | ₽0.86/изобр. |
| FLUX-1-Redux-dev | black forest labs | ₽1.03/изобр. |
| Flux 2 Klein 4b | black forest labs | ₽1.20/изобр. |
| Flux 2 Klein 9b | black forest labs | ₽1.28/изобр. |
| Flux 2 Pro | black forest labs | ₽1.28/изобр. |
| bria remove background | bria remove background | ₽1.54/изобр. |
| Qwen Image Edit | alibaba | ₽2.14/изобр. |
| Wan2.6-T2I | Wan AI | ₽2.57/изобр. |
| blur_background | Bria | ₽3.42/изобр. |
| Bria-3.2 | Bria | ₽3.42/изобр. |
| Bria-3.2-vector | Bria | ₽3.42/изобр. |
| erase_foreground | Bria | ₽3.42/изобр. |
| expand | Bria | ₽3.42/изобр. |
| fibo | Bria | ₽3.42/изобр. |
| fibo_edit | Bria | ₽3.42/изобр. |
| Flux 1.1 Pro | FLUX 1.1 pro | ₽3.42/изобр. |
| Seedream-4 | ByteDance | ₽3.42/изобр. |
| Qwen Image Max | alibaba | ₽6.42/изобр. |
| FLUX.2 [max] | black forest labs | ₽8.56/изобр. |
Видео и аудио
за секунду · 7
▼
Видео и аудио
за секунду · 7
| Модель | Maker | Цена |
|---|---|---|
| nemotron-3.5-asr-streaming-multilingual-0.6b | nemotron 3 5 asr streaming multilingual 0 6b | ₽0.0003/сек |
| qwen3-asr-0.6b | qwen3 asr 0 6b | ₽0.0003/сек |
| Whisper Large V3 Turbo | openai | ₽0.0003/сек |
| qwen3-asr-1.7b | qwen3 asr 1 7b | ₽0.0006/сек |
| Whisper Large V3 | whisper large v3 | ₽0.0006/сек |
| Voxtral-Mini-3B-2507 | mistralai | ₽0.0014/сек |
| Voxtral Small 24B 2507 | mistralai | ₽0.0043/сек |
Озвучка
за 1M символов · 11
▼
Озвучка
за 1M символов · 11
| Модель | Maker | Цена |
|---|---|---|
| Kokoro-82M | hexgrad | ₽53.07/1M симв. |
| chatterbox-multilingual | ResembleAI | ₽85.60/1M симв. |
| chatterbox-turbo | ResembleAI | ₽85.60/1M симв. |
| Audio8-TTS-Preview-0.6b | Audio8 | ₽428/1M симв. |
| csm-1b | sesame | ₽599/1M симв. |
| HiggsAudioV2.5 | bosonai | ₽1 712/1M симв. |
| Qwen3-TTS | Qwen | ₽1 712/1M симв. |
| Qwen3-TTS-VoiceDesign | Qwen | ₽1 712/1M симв. |
| realtime-tts-1.5-mini | inworld ai | ₽2 140/1M симв. |
| realtime-tts-2 | inworld ai | ₽2 996/1M симв. |
| realtime-tts-1.5-max | inworld ai | ₽4 280/1M симв. |
Без публичной цены
нет данных в открытых источниках · 15
▼
Без публичной цены
нет данных в открытых источниках · 15
| Модель | Maker | Примечание |
|---|---|---|
| Cosmos3-Nano | nvidia | — |
| Cosmos3-Super | nvidia | — |
| gemini 3 pro image | — | |
| MiMo-V2.5-tts | XiaomiMiMo | — |
| MiMo-V2.5-tts-voicedesign | XiaomiMiMo | — |
| Nano Banana 2 Lite | — | |
| p-video | PrunaAI | — |
| Pixverse-6-T2V | Pixverse | — |
| Pixverse-T2V | Pixverse | — |
| Pixverse-T2V-HD | Pixverse | — |
| Seedance 1.5 Pro | bytedance | — |
| Seedance 2.0 | bytedance | — |
| Veo 3.1 | — | |
| Wan2.2-T2V-A14B | Wan AI | — |
| Wan2.6-T2V | Wan AI | — |
Текст и чат
оплата за токены · 125
▼
Текст и чат
оплата за токены · 125
| Модель | Maker | Input | Output | TCO/мес |
|---|---|---|---|---|
| Embeddinggemma 300m | ₽0.17/M | ₽0.17/M | ₽3 | |
| all-MiniLM-L12-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| all-MiniLM-L6-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| all-mpnet-base-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| bge-base-en-v1.5 | baai | ₽0.43/M | ₽0.43/M | ₽9 |
| clip-ViT-B-32 | sentence transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| clip-ViT-B-32-multilingual-v1 | sentence transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| E5-Base-v2 | Intfloat | ₽0.43/M | ₽0.43/M | ₽9 |
| GTE-Base | Thenlper | ₽0.43/M | ₽0.43/M | ₽9 |
| multi-qa-mpnet-base-dot-v1 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| paraphrase-MiniLM-L6-v2 | Sentence Transformers | ₽0.43/M | ₽0.43/M | ₽9 |
| text2vec-base-chinese | shibing624 | ₽0.43/M | ₽0.43/M | ₽9 |
| Bge En Icl | BAAI | ₽0.86/M | ₽0.86/M | ₽17 |
| Bge M3 | baai | ₽0.86/M | ₽0.86/M | ₽17 |
| Bge M3 Multi | BAAI | ₽0.86/M | ₽0.86/M | ₽17 |
| bge-large-en-v1.5 | baai | ₽0.86/M | ₽0.86/M | ₽17 |
| E5-Large-v2 | Intfloat | ₽0.86/M | ₽0.86/M | ₽17 |
| GTE-Large | Thenlper | ₽0.86/M | ₽0.86/M | ₽17 |
| Llama Nemotron Embed Vl 1b V2 | nvidia | ₽0.86/M | ₽0.86/M | ₽17 |
| Multilingual-E5-Large | Intfloat | ₽0.86/M | ₽0.86/M | ₽17 |
| Qwen3 Embedding 8B | qwen3 embedding 8b | ₽0.86/M | ₽0.86/M | ₽17 |
| qwen3-embedding-0.6b | qwen | ₽0.86/M | ₽0.86/M | ₽17 |
| Qwen3 Embedding 4B | qwen3 embedding 4b | ₽1.71/M | ₽1.71/M | ₽34 |
| Mistral-Nemo-Instruct-2407 | mistralai | ₽1.63/M | ₽2.57/M | ₽45 |
| Meta-Llama-3.1-8B-Instruct-Turbo | meta llama | ₽1.71/M | ₽3.42/M | ₽51 |
| gemma-4-E4B-it | ₽1.71/M | ₽8.56/M | ₽77 | |
| L3-8B-Lunaris-v1-Turbo | Sao10K | ₽3.42/M | ₽4.28/M | ₽90 |
| granite-4.2-3b | ibm granite | ₽2.57/M | ₽10.27/M | ₽103 |
| gpt oss 20b | openai | ₽2.57/M | ₽11.98/M | ₽111 |
| mistralai · mistral-small-24b-instruct-2501 | mistralai | ₽4.28/M | ₽6.85/M | ₽120 |
| Gemma 3 4B It | gemma 3 4b it | ₽4.28/M | ₽8.56/M | ₽128 |
| gpt oss 120b | openai | ₽3.17/M | ₽14.55/M | ₽136 |
| gemma 3 12b it | ₽4.28/M | ₽12.84/M | ₽150 | |
| Nvidia Nemotron 3 Nano 30B A3B | nvidia nemotron 3 nano 30b a3b | ₽4.28/M | ₽17.12/M | ₽171 |
| Ling-3.0-flash | inclusionai | ₽5.14/M | ₽15.41/M | ₽180 |
| Phi 4 | Microsoft | ₽5.99/M | ₽11.98/M | ₽180 |
| gemma 3 27b it | ₽6.85/M | ₽13.70/M | ₽205 | |
| granite-4.2-8b | ibm granite | ₽5.14/M | ₽21.40/M | ₽210 |
| deepseek v4 flash | deepseek | ₽6.85/M | ₽15.41/M | ₽214 |
| Mistral-Small-3.2-24B-Instruct-2506 | mistralai | ₽6.42/M | ₽17.12/M | ₽214 |
| Nemotron 3.5 Lightning | NVIDIA | ₽6.85/M | ₽17.12/M | ₽223 |
| qwen3.5 9b | qwen | ₽8.56/M | ₽12.84/M | ₽235 |
| qwen3 32b | qwen | ₽6.85/M | ₽23.97/M | ₽257 |
| gemma 4 26b a4b it | ₽5.99/M | ₽29.10/M | ₽265 | |
| glm 4.7 flash | z ai | ₽5.14/M | ₽34.24/M | ₽274 |
| gemma-4-31B-it-turbo | ₽7.70/M | ₽29.10/M | ₽300 | |
| Llama 4 Scout 17b 16e Instruct | meta | ₽8.56/M | ₽25.68/M | ₽300 |
| Llama-3.3-70B-Instruct-Turbo | meta llama | ₽8.56/M | ₽27.39/M | ₽308 |
| qwen3 14b | qwen | ₽10.27/M | ₽20.54/M | ₽308 |
| Nvidia Nemotron 3 Super 120B A12B | nvidia nemotron 3 super 120b a12b | ₽7.28/M | ₽34.24/M | ₽317 |
| Dola-Seed 2.0 Mini | bytedance | ₽8.56/M | ₽34.24/M | ₽342 |
| llama guard 4 12b | meta llama | ₽15.41/M | ₽15.41/M | ₽385 |
| gemma 4 31b it | ₽11.13/M | ₽32.53/M | ₽385 | |
| Qwen3 235B A22B Instruct | alibaba | ₽7.70/M | ₽47.08/M | ₽389 |
| qwen3 30b a3b | qwen | ₽10.27/M | ₽42.80/M | ₽419 |
| Nemotron-Content-Safety-3.5 | nvidia | ₽17.12/M | ₽17.12/M | ₽428 |
| glm 5.3 flash | z ai | ₽12.84/M | ₽42.80/M | ₽471 |
| Hy3 | hy3 | ₽11.98/M | ₽49.65/M | ₽488 |
| gpt-oss-120b-Turbo | openai | ₽12.84/M | ₽51.36/M | ₽514 |
| Qwen3 VL 30B | qwen3 vl 30b a3b | ₽12.84/M | ₽51.36/M | ₽514 |
| granite-4.2-30b | ibm granite | ₽13.70/M | ₽55.64/M | ₽552 |
| qwen3.6 35b a3b | qwen | ₽8.56/M | ₽81.32/M | ₽578 |
| deepseek v3.2 | deepseek | ₽22.26/M | ₽32.53/M | ₽608 |
| qwen3 next 80b a3b | qwen3 next 80b a3b | ₽7.70/M | ₽94.16/M | ₽625 |
| qwen3.5 35b a3b | qwen | ₽11.98/M | ₽85.60/M | ₽668 |
| Llama-4-Maverick-17B-128E-Instruct-FP8 | meta llama | ₽17.12/M | ₽68.48/M | ₽685 |
| qwen3 vl 235b a22b | qwen3 vl 235b a22b | ₽17.12/M | ₽75.33/M | ₽719 |
| gpt-oss-120b-Ultra | openai | ₽17.12/M | ₽81.32/M | ₽749 |
| gemma-4-31B-it-Ultra | ₽23.11/M | ₽65.06/M | ₽788 | |
| qwen 2.5 72b instruct | qwen | ₽30.82/M | ₽34.24/M | ₽788 |
| DeepSeek-V3 | deepseek ai | ₽20.54/M | ₽77.04/M | ₽796 |
| Step 3.7 Flash | step 3.7 flash | ₽17.12/M | ₽98.44/M | ₽835 |
| DeepSeek V3.1 | deepseek ai | ₽21.40/M | ₽81.32/M | ₽835 |
| Meta-Llama-3.1-70B-Instruct-Turbo | meta llama | ₽34.24/M | ₽34.24/M | ₽856 |
| minimax m2.7 | minimax | ₽21.40/M | ₽85.60/M | ₽856 |
| Mythomax L2 13B | Gryphe | ₽34.24/M | ₽34.24/M | ₽856 |
| Qwen3-Coder-480B-A35B-Instruct-Turbo | Qwen | ₽25.68/M | ₽85.60/M | ₽942 |
| minimax m3 | minimax | ₽23.97/M | ₽94.16/M | ₽950 |
| Muse Glimmer 30B | muse glimmer 30b | ₽25.68/M | ₽103/M | ₽1 027 |
| gemini 3.1 flash lite | ₽21.40/M | ₽128/M | ₽1 070 | |
| ByteDance Seed 1.8 | bytedance | ₽21.40/M | ₽171/M | ₽1 284 |
| Inkling Small | thinkingmachines | ₽38.52/M | ₽103/M | ₽1 284 |
| MiniMax-M2.7-Turbo | MiniMaxAI | ₽32.53/M | ₽146/M | ₽1 378 |
| glm 4.7 | z ai | ₽34.24/M | ₽150/M | ₽1 434 |
| Hermes 3 70B Instruct | nousresearch | ₽59.92/M | ₽59.92/M | ₽1 498 |
| qwen3.5 122b a10b | qwen | ₽24.82/M | ₽205/M | ₽1 524 |
| mimo v2.5 | mimo v2.5 | ₽34.24/M | ₽171/M | ₽1 541 |
| qwen3.5 27b | qwen | ₽22.26/M | ₽223/M | ₽1 558 |
| Gemini 2.5 Flash | gemini 2.5 flash | ₽25.68/M | ₽214/M | ₽1 584 |
| glm 4.6 | z ai | ₽42.80/M | ₽171/M | ₽1 712 |
| Kimi K2.5 | kimi k2.5 | ₽38.52/M | ₽193/M | ₽1 733 |
| DeepSeek R1 | deepseek r1 | ₽42.80/M | ₽184/M | ₽1 776 |
| NVIDIA-Nemotron-3-Ultra-550B-A55B | nvidia | ₽42.80/M | ₽188/M | ₽1 798 |
| L3.1-70B-Euryale-v2.2 | Sao10K | ₽72.76/M | ₽72.76/M | ₽1 819 |
| GLM-5 | glm 5 | ₽51.36/M | ₽178/M | ₽1 917 |
| qwen3.6 27b | qwen | ₽27.39/M | ₽274/M | ₽1 917 |
| qwen3.8 27b | qwen | ₽34.24/M | ₽257/M | ₽1 969 |
| qwen3.5 397b a17b | qwen | ₽38.52/M | ₽257/M | ₽2 054 |
| Dola-Seed 2.0 Pro | bytedance | ₽42.80/M | ₽257/M | ₽2 140 |
| seed-2.0-code | seed 2 0 code | ₽42.80/M | ₽257/M | ₽2 140 |
| Hermes 3 405B Instruct | nousresearch | ₽85.60/M | ₽85.60/M | ₽2 140 |
| glm 5.2 | z ai | ₽64.20/M | ₽205/M | ₽2 311 |
| kimi k2.7 code | moonshotai | ₽58.21/M | ₽291/M | ₽2 619 |
| kimi k2.6 | moonshotai | ₽64.20/M | ₽300/M | ₽2 782 |
| gemini 3.7 flash | ₽64.20/M | ₽321/M | ₽2 889 | |
| mimo v2.5 pro | mimo v2.5 pro | ₽85.60/M | ₽257/M | ₽2 996 |
| glm 5.1 | z ai | ₽89.88/M | ₽300/M | ₽3 296 |
| deepseek v4 pro | deepseek | ₽111/M | ₽223/M | ₽3 338 |
| Inkling | Thinking Machines | ₽81.32/M | ₽347/M | ₽3 360 |
| glm 5.3 | z ai | ₽103/M | ₽342/M | ₽3 766 |
| Claude Haiku 4.5 | claude haiku 4 5 | ₽85.60/M | ₽428/M | ₽3 852 |
| Qwen3 Max | qwen3 max | ₽103/M | ₽514/M | ₽4 622 |
| qwen3.8 max | qwen | ₽141/M | ₽424/M | ₽4 944 |
| Qwen3.8 2.4t A95B | qwen3.8 2.4t a95b | ₽171/M | ₽514/M | ₽5 992 |
| gemini 3.5 flash | ₽128/M | ₽770/M | ₽6 420 | |
| Gemini 2.5 Pro | gemini 2.5 pro | ₽107/M | ₽856/M | ₽6 420 |
| qwen3.7 max | qwen | ₽214/M | ₽642/M | ₽7 490 |
| claude sonnet 5 | anthropic | ₽171/M | ₽856/M | ₽7 704 |
| gemini 3.1 pro | ₽171/M | ₽1 027/M | ₽8 560 | |
| kimi k3 | moonshotai | ₽244/M | ₽1 220/M | ₽10 978 |
| Claude Sonnet 4.6 | claude sonnet 4 6 | ₽257/M | ₽1 284/M | ₽11 556 |
| Claude Opus 4.7 | claude opus 4 7 | ₽428/M | ₽2 140/M | ₽19 260 |
| claude opus 4.8 | anthropic | ₽428/M | ₽2 140/M | ₽19 260 |
| claude opus 5 | anthropic | ₽428/M | ₽2 140/M | ₽19 260 |
| claude fable 5 | anthropic | ₽856/M | ₽4 280/M | ₽38 520 |
Изображения
за изображение · 24
▼
Изображения
за изображение · 24
| Модель | Maker | Цена |
|---|---|---|
| Sdxl Turbo | stabilityai | ₽0.02/изобр. |
| Flux 1 Schnell | black forest labs | ₽0.04/изобр. |
| p-image | PrunaAI | ₽0.43/изобр. |
| FLUX.1 [dev] | black forest labs | ₽0.77/изобр. |
| Flux 2 Dev | black forest labs | ₽0.86/изобр. |
| FLUX.1-Kontext-dev | black forest labs | ₽0.86/изобр. |
| FLUX-1-Redux-dev | black forest labs | ₽1.03/изобр. |
| Flux 2 Klein 4b | black forest labs | ₽1.20/изобр. |
| Flux 2 Klein 9b | black forest labs | ₽1.28/изобр. |
| Flux 2 Pro | black forest labs | ₽1.28/изобр. |
| bria remove background | bria remove background | ₽1.54/изобр. |
| Qwen Image Edit | alibaba | ₽2.14/изобр. |
| Wan2.6-T2I | Wan AI | ₽2.57/изобр. |
| blur_background | Bria | ₽3.42/изобр. |
| Bria-3.2 | Bria | ₽3.42/изобр. |
| Bria-3.2-vector | Bria | ₽3.42/изобр. |
| erase_foreground | Bria | ₽3.42/изобр. |
| expand | Bria | ₽3.42/изобр. |
| fibo | Bria | ₽3.42/изобр. |
| fibo_edit | Bria | ₽3.42/изобр. |
| Flux 1.1 Pro | FLUX 1.1 pro | ₽3.42/изобр. |
| Seedream-4 | ByteDance | ₽3.42/изобр. |
| Qwen Image Max | alibaba | ₽6.42/изобр. |
| FLUX.2 [max] | black forest labs | ₽8.56/изобр. |
Видео и аудио
за секунду · 7
▼
Видео и аудио
за секунду · 7
| Модель | Maker | Цена |
|---|---|---|
| nemotron-3.5-asr-streaming-multilingual-0.6b | nemotron 3 5 asr streaming multilingual 0 6b | ₽0.0003/сек |
| qwen3-asr-0.6b | qwen3 asr 0 6b | ₽0.0003/сек |
| Whisper Large V3 Turbo | openai | ₽0.0003/сек |
| qwen3-asr-1.7b | qwen3 asr 1 7b | ₽0.0006/сек |
| Whisper Large V3 | whisper large v3 | ₽0.0006/сек |
| Voxtral-Mini-3B-2507 | mistralai | ₽0.0014/сек |
| Voxtral Small 24B 2507 | mistralai | ₽0.0043/сек |
Озвучка
за 1M символов · 11
▼
Озвучка
за 1M символов · 11
| Модель | Maker | Цена |
|---|---|---|
| Kokoro-82M | hexgrad | ₽53.07/1M симв. |
| chatterbox-multilingual | ResembleAI | ₽85.60/1M симв. |
| chatterbox-turbo | ResembleAI | ₽85.60/1M симв. |
| Audio8-TTS-Preview-0.6b | Audio8 | ₽428/1M симв. |
| csm-1b | sesame | ₽599/1M симв. |
| HiggsAudioV2.5 | bosonai | ₽1 712/1M симв. |
| Qwen3-TTS | Qwen | ₽1 712/1M симв. |
| Qwen3-TTS-VoiceDesign | Qwen | ₽1 712/1M симв. |
| realtime-tts-1.5-mini | inworld ai | ₽2 140/1M симв. |
| realtime-tts-2 | inworld ai | ₽2 996/1M симв. |
| realtime-tts-1.5-max | inworld ai | ₽4 280/1M симв. |
Без публичной цены
нет данных в открытых источниках · 15
▼
Без публичной цены
нет данных в открытых источниках · 15
| Модель | Maker | Примечание |
|---|---|---|
| Cosmos3-Nano | nvidia | — |
| Cosmos3-Super | nvidia | — |
| gemini 3 pro image | — | |
| MiMo-V2.5-tts | XiaomiMiMo | — |
| MiMo-V2.5-tts-voicedesign | XiaomiMiMo | — |
| Nano Banana 2 Lite | — | |
| p-video | PrunaAI | — |
| Pixverse-6-T2V | Pixverse | — |
| Pixverse-T2V | Pixverse | — |
| Pixverse-T2V-HD | Pixverse | — |
| Seedance 1.5 Pro | bytedance | — |
| Seedance 2.0 | bytedance | — |
| Veo 3.1 | — | |
| Wan2.2-T2V-A14B | Wan AI | — |
| Wan2.6-T2V | Wan AI | — |
О провайдере
О сервисе
DeepInfra — direct inference platform (не multi-vendor router): модели деплоятся на GPU DeepInfra, billing pay-per-use. Конкурирует с Together, Fireworks, SiliconFlow на open-weight моделях. Часто дешевле USD-агрегаторов на Llama/Qwen/FLUX. На AI-APISHKA — flat $/1M из metadata.pricing без ratio-пересчёта.
Как сравнивать на AI-APISHKA
Сравнивайте $/1M напрямую с Together/Fireworks/OpenRouter. Image — per_image_unit ($/image). Для embeddings — только input_tokens. Prompt cache — cache_read_tokens.
Частые вопросы
- DeepInfra — агрегатор или inference platform?
- Inference platform: модели крутятся на GPU DeepInfra, не прокси к OpenAI/Anthropic (хотя есть и проприетарные модели).
- Где цены DeepInfra?
- GET /v1/openai/models → metadata.pricing: input_tokens, output_tokens ($/1M), per_image_unit, input_seconds, input_characters.
- DeepInfra vs SiliconFlow?
- Оба direct inference. DeepInfra — USD, SiliconFlow — CNY. Сравнивайте $/1M на overlap-моделях (Qwen, DeepSeek, FLUX).
Способы оплаты и доступные валюты могут меняться — актуальная информация на сайте провайдера.
Цена не сходится или нашли ошибку?
DeepInfra предлагает 183 моделей AI API. Средняя цена — ₽84/M (25M токенов/мес · ввод 80% / вывод 20%). Самая низкая цена input — ₽0.17/M (Embeddinggemma 300m).
Сравните цены на nemotron-3.5-asr-streaming-multilingual-0.6b , qwen3-asr-0.6b , Whisper Large V3 Turbo , qwen3-asr-1.7b и Whisper Large V3 у других провайдеров или в каталоге агрегаторов . Данные обновлены 30.08.2026.