# Runcrate pricing

Served 2026-11-02T02:17:51.684Z from Runcrate's live catalog, the same data as https://www.runcrate.ai/pricing and https://www.runcrate.ai/inference/pricing. The time GPU stock was read is under GPUs. Stock and prices change through the day, so read this page again before you quote a price.

- GPUs as JSON (memory, price, stock, regions, prepaid terms; a GPU sold out in the last 30 days stays listed at `available_options: 0`): https://www.runcrate.ai/gpus.json
- Models as JSON: https://www.runcrate.ai/api/models/catalog
- Act on it from an agent: the MCP server https://mcp.runcrate.ai/mcp (OAuth, no key), or the CLI: `curl -sSL https://runcrate.ai/install.sh | sh`
- Everything else an agent needs: https://www.runcrate.ai/llms.txt

Pay as you go from prepaid credits: no commitment and no minimum, and credits never expire. No quota requests: the listed rate is the rate, for any GPU in stock. Top up without a credit card: the dashboard (Billing, then Add credits) takes other ways to pay.

## GPUs on demand

Stock and prices read 2026-11-02T02:15:20.407Z. Prices per GPU-hour start at the figure shown for what is in stock now, billed by the second unless the row says by the minute. No commitment: stop the machine and billing stops.

| GPU | Memory | From per GPU-hour | GPUs per machine | Offers listed | Billed |
| --- | --- | --- | --- | --- | --- |
| B300 | 288GB HBM3e | $11.57 | 1, 8 | 6 | by the second |
| B200 | 180GB HBM3e | $7.67 | 1, 2, 8 | 5 | by the second |
| H200 | 141GB HBM3e | $4.63 | 1, 2, 4, 8 | 13 | by the minute |
| H100 | 80GB HBM3 | $2.49 | 1, 2, 4 | 12 | by the second |
| GH200 | 96GB HBM3 | $2.98 | 1 | 1 | by the second |
| A100 80GB | 80GB HBM2e | $1.73 | 1, 2, 8 | 22 | by the second |
| A100 40GB | 40GB HBM2 | $1.99 | 1, 8 | 8 | by the second |
| L40S | 48GB GDDR6 | $0.54 | 1, 2, 4, 8 | 19 | by the second |
| L40 | 48GB GDDR6 | $1.14 | 1, 2, 4 | 12 | by the second |
| RTX 5090 | 32GB GDDR7 | $0.84 | 1, 2, 4 | 16 | by the second |
| RTX 4090 | 24GB GDDR6X | $0.57 | 1, 2, 8 | 4 | by the second |
| RTX 6000 Ada | 48GB GDDR6 | $1.18 | 1 | 7 | by the second |
| RTX PRO 6000 | 96GB GDDR7 | $2.22 | 1, 2, 4, 8 | 19 | by the second |
| A6000 | 48GB GDDR6 | $0.57 | 1, 2, 4, 8 | 36 | by the second |
| A5000 | 24GB GDDR6 | $1.85 | 1, 2, 4 | 3 | by the second |
| L4 | 24GB GDDR6 | $1.24 | 1, 2, 4 | 4 | by the second |
| A4000 | 16GB GDDR6 | $1.04 | 1, 2, 4 | 3 | by the second |
| RTX 4000 Ada | 20GB GDDR6 | $1.03 | 1 | 1 | by the second |
| A10 | 24GB GDDR6 | $1.68 | 1 | 2 | by the second |
| V100 32GB | 32GB HBM2 | $0.36 | 1, 2, 4, 8 | 11 | by the second |
| V100 16GB | 16GB HBM2 | $0.36 | 1, 2, 4 | 5 | by the second |

### Prepaid terms

Paid upfront for the whole machine and the whole term, and not cancellable. Shown for each length that costs less per GPU-hour than on demand.

| GPU | Months | Per GPU-hour | Off on demand | Per GPU for the term |
| --- | --- | --- | --- | --- |
| H200 | 1 | $4.3985 | 5% | $3210.91 |
| H200 | 3 | $4.2596 | 8% | $9328.52 |
| H200 | 6 | $4.1207 | 11% | $18048.67 |
| H200 | 12 | $3.9355 | 15% | $34474.98 |
| GH200 | 1 | $2.9055 | 2.5% | $2121.02 |
| GH200 | 3 | $2.831 | 5% | $6199.89 |
| GH200 | 6 | $2.7416 | 8% | $12008.21 |
| GH200 | 12 | $2.6522 | 11% | $23233.27 |
| A100 80GB | 3 | $1.6867 | 2.5% | $3693.98 |
| A100 80GB | 6 | $1.5916 | 8% | $6971.21 |
| A100 80GB | 12 | $1.5397 | 11% | $13487.77 |
| A100 40GB | 1 | $1.8905 | 5% | $1380.07 |
| A100 40GB | 3 | $1.8308 | 8% | $4009.45 |
| A100 40GB | 6 | $1.7711 | 11% | $7757.42 |
| A100 40GB | 12 | $1.6915 | 15% | $14817.54 |
| L40 | 1 | $1.1115 | 2.5% | $811.40 |
| L40 | 3 | $1.083 | 5% | $2371.77 |
| L40 | 6 | $1.0488 | 8% | $4593.74 |
| L40 | 12 | $1.0146 | 11% | $8887.90 |
| RTX 4090 | 1 | $0.5415 | 5% | $395.30 |
| RTX 4090 | 3 | $0.5244 | 8% | $1148.44 |
| RTX 4090 | 6 | $0.5073 | 11% | $2221.97 |
| RTX 4090 | 12 | $0.4845 | 15% | $4244.22 |
| A5000 | 1 | $1.8038 | 2.5% | $1316.74 |
| A5000 | 3 | $1.7575 | 5% | $3848.93 |
| A5000 | 6 | $1.702 | 8% | $7454.76 |
| A5000 | 12 | $1.6465 | 11% | $14423.34 |
| L4 | 1 | $1.209 | 2.5% | $882.57 |
| L4 | 3 | $1.178 | 5% | $2579.82 |
| L4 | 6 | $1.1408 | 8% | $4996.70 |
| L4 | 12 | $1.1036 | 11% | $9667.54 |
| A4000 | 1 | $1.014 | 2.5% | $740.22 |
| A4000 | 3 | $0.988 | 5% | $2163.72 |
| A4000 | 6 | $0.9568 | 8% | $4190.78 |
| A4000 | 12 | $0.9256 | 11% | $8108.26 |
| RTX 4000 Ada | 1 | $1.0042 | 2.5% | $733.10 |
| RTX 4000 Ada | 3 | $0.9785 | 5% | $2142.92 |
| RTX 4000 Ada | 6 | $0.9476 | 8% | $4150.49 |
| RTX 4000 Ada | 12 | $0.9167 | 11% | $8030.29 |
| A10 | 1 | $1.638 | 2.5% | $1195.74 |
| A10 | 3 | $1.596 | 5% | $3495.24 |
| A10 | 6 | $1.5456 | 8% | $6769.73 |
| A10 | 12 | $1.4952 | 11% | $13097.95 |

## Inference

Billed prices in USD, the exact rate each request is charged. OpenAI-compatible at https://api.runcrate.ai/v1. A model marked free tier is free with GitHub linked, then billed at the price shown (https://www.runcrate.ai/inference/free).

### Text generation

| Model | Input per 1M tokens | Output per 1M tokens | Cached input per 1M tokens |
| --- | --- | --- | --- |
| `cohere/command-a-plus` | $0.315 | $1.575 | none |
| `deepseek-ai/DeepSeek-R1-0528` | $0.525 | $2.2575 | $0.3675 |
| `deepseek-ai/DeepSeek-V3.2` | $0.273 | $0.399 | $0.1365 |
| `deepseek-ai/DeepSeek-V4-Flash` | $0.0945 | $0.189 | $0.0189 |
| `deepseek-ai/DeepSeek-V4-Flash-0731` | $0.063 | $0.189 | $0.01575 |
| `deepseek-ai/DeepSeek-V4-Pro` | $1.365 | $2.73 | $0.105 |
| `deepseek-ai/DeepSeek-V4-Pro-0813` | $1.365 | $2.73 | $0.105 |
| `deepseek-ai/DeepSeek-V4.1-Flash` (free tier) | $0.0735 ($0.147 when busy) | $0.2205 ($0.441 when busy) | $0.002205 ($0.00441 when busy) |
| `deepseek-ai/DeepSeek-V4.1-Flash-1017` (free tier) | $0.05145 ($0.1029 when busy) | $0.15435 ($0.3087 when busy) | $0.001029 ($0.002058 when busy) |
| `google/gemma-3-12b-it` | $0.0525 | $0.1575 | none |
| `google/gemma-3-27b-it` | $0.084 | $0.168 | none |
| `google/gemma-3-4b-it` | $0.0525 | $0.105 | none |
| `google/gemma-4-26B-A4B-it` | $0.0735 | $0.357 | none |
| `google/gemma-4-31B-it` | $0.21 | $0.42 | none |
| `google/gemma-4-31B-it-Ultra` | $0.2835 | $0.798 | none |
| `google/gemma-4-e2b` | $0.042 | $0.084 | none |
| `ibm-granite/granite-4.2-8b` | $0.063 | $0.2625 | $0.01575 |
| `inclusionAI/Ling-3.1-flash` | $0.07875 | $0.23625 | $0.01575 |
| `LiquidAI/lfm-2.5-24b-a2b` | $0.0525 | $0.735 | none |
| `meta-llama/Llama-3.3-70B-Instruct-Turbo` | $0.105 | $0.336 | none |
| `meta-llama/Llama-4-Scout-17B-16E-Instruct` | $0.105 | $0.315 | none |
| `meta-llama/Llama-Guard-4-12B` | $0.189 | $0.189 | none |
| `meta-llama/muse-spark-open` | $1.3125 | $4.4625 | $0.1575 |
| `meta-models/Muse-Glimmer-30B` | $0.315 | $1.26 | $0.042 |
| `microsoft/phi-4` | $0.0735 | $0.147 | none |
| `MiniMaxAI/MiniMax-M3` | $0.294 | $1.155 | $0.0588 |
| `MiniMaxAI/minimax-m3-1-flash` | $0.1575 | $0.63 | $0.01575 |
| `mistralai/devstral-2-123b` | $0.42 | $2.10 | $0.042 |
| `mistralai/magistral-small-2509` | $0.525 | $1.575 | $0.07875 |
| `mistralai/ministral-3-14b-instruct` | $0.21 | $0.21 | $0.021 |
| `mistralai/ministral-3-3b-instruct` | $0.105 | $0.105 | $0.0105 |
| `mistralai/ministral-3-8b-instruct` | $0.1575 | $0.1575 | $0.01575 |
| `mistralai/mistral-large-3` | $0.525 | $1.575 | $0.0525 |
| `mistralai/Mistral-Nemo-Instruct-2407` | $0.03045 | $0.0315 | none |
| `mistralai/Mistral-Small-3.2-24B-Instruct-2506` | $0.07875 | $0.21 | none |
| `mistralai/mistral-small-4` | $0.1575 | $0.63 | none |
| `moondream/moondream3-preview` | $0.315 | $2.625 | none |
| `moondream/moondream3.1-9b-a2b` | $0.315 | $2.625 | none |
| `moonshotai/Kimi-K2.6` | $0.7875 | $3.675 | $0.1575 |
| `moonshotai/Kimi-K3` | $2.9925 | $14.9625 | $0.29925 |
| `nvidia/Nemotron-Content-Safety-3.5` | $0.21 | $0.21 | none |
| `nvidia/NVIDIA-Nemotron-3-Super-120B-A12B` | $0.08925 | $0.42 | none |
| `nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B` | $0.525 | $2.31 | $0.105 |
| `nvidia/NVIDIA-Nemotron-3.5-Lightning` | $0.063 | $0.168 | $0.0315 |
| `nvidia/nvidia-nemotron-nano-12b-v2` | $0.21 | $0.63 | none |
| `openai/gpt-oss-120b` (free tier) | $0.019425 ($0.03885 when busy) | $0.08925 ($0.1785 when busy) | none |
| `openai/gpt-oss-120b-Ultra` | $0.21 | $0.9975 | none |
| `openai/gpt-oss-20b` | $0.0315 | $0.147 | $0.01575 |
| `openai/gpt-oss-safeguard-120b` | $0.1575 | $0.63 | none |
| `Qwen/Qwen3-235B-A22B-Instruct-2507` | $0.0945 | $0.5775 | none |
| `Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo` | $0.315 | $1.05 | $0.105 |
| `Qwen/Qwen3-Next-80B-A3B-Instruct` | $0.0945 | $1.155 | none |
| `Qwen/Qwen3.5-27B` | $0.273 | $2.73 | none |
| `Qwen/Qwen3.5-35B-A3B` | $0.147 | $1.05 | $0.0525 |
| `Qwen/Qwen3.5-397B-A17B` | $0.4725 | $3.15 | $0.231 |
| `Qwen/Qwen3.5-9B` | $0.105 | $0.1575 | none |
| `Qwen/Qwen3.6-27B` | $0.336 | $3.36 | none |
| `Qwen/Qwen3.6-35B-A3B` | $0.105 | $0.9975 | $0.01575 |
| `Qwen/Qwen3.8-27B` | $0.1575 | $1.96875 | $0.039375 |
| `Qwen/Qwen3.8-Flash` (free tier) | $0.059325 ($0.11865 when busy) | $0.20055 ($0.4011 when busy) | $0.006327825 ($0.01265565 when busy) |
| `reflection/beam-1.0-501b-a23b` | $0.2625 | $1.05 | $0.0525 |
| `tencent/Hy4-preview` | $0.8757 | $2.62605 | $0.0441 |
| `thinkingmachines/Inkling-Small` | $0.4725 | $1.26 | $0.105 |
| `XiaomiMiMo/MiMo-V2.6-Flash` | $0.147 | $0.294 | $0.00294 |
| `XiaomiMiMo/MiMo-V2.6-Pro` | $0.40635 | $0.82215 | $0.003402 |
| `zai-org/GLM-4.7` | $0.42 | $1.8375 | $0.084 |
| `zai-org/GLM-5.2` | $0.590625 | $1.89 | $0.11025 |
| `zai-org/GLM-5.3` (free tier) | $0.2953125 ($0.590625 when busy) | $1.3125 ($2.625 when busy) | $0.065625 ($0.13125 when busy) |
| `zai-org/GLM-5.3-Flash` (free tier) | $0.07875 ($0.1575 when busy) | $0.2625 ($0.525 when busy) | $0.01575 ($0.0315 when busy) |

### Vision & multimodal

| Model | Input per 1M tokens | Output per 1M tokens | Cached input per 1M tokens |
| --- | --- | --- | --- |
| `deepseek-ai/DeepSeek-V4-Flash-Vision-Exp` | $0.22638 | $0.67914 | $0.007203 |
| `inclusionAI/Ling-3.0-flash-VL` | $0.063 | $0.189 | $0.0126 |
| `nvidia/NVIDIA-Nemotron-3.5-Super-VL` | $0.1575 | $0.525 | $0.021 |
| `Qwen/Qwen3-VL-235B-A22B-Instruct` | $0.21 | $0.924 | $0.1155 |
| `Qwen/Qwen3-VL-30B-A3B-Instruct` | $0.1575 | $0.63 | none |

### Image generation

| Model | Price | Per |
| --- | --- | --- |
| `black-forest-labs/FLUX-2-klein-4b` | $0.00105 | megapixel |
| `black-forest-labs/FLUX.1-dev` | $0.02625 | image |
| `black-forest-labs/FLUX.1-kontext-dev` | $0.02625 | image |
| `black-forest-labs/FLUX.1-kontext-max` | $0.084 | image |
| `black-forest-labs/FLUX.1-kontext-pro` | $0.042 | image |
| `black-forest-labs/FLUX.1-krea-dev` | $0.042 | image |
| `black-forest-labs/FLUX.1-schnell` | $0.00315 | image |
| `black-forest-labs/FLUX.1.1-pro` | $0.042 | image |
| `black-forest-labs/FLUX.2-dev` | $0.0126 | megapixel |
| `black-forest-labs/FLUX.2-flex` | $0.063 | megapixel |
| `black-forest-labs/FLUX.2-max` | $0.0735 | megapixel |
| `black-forest-labs/FLUX.2-pro` | $0.0315 | megapixel |
| `ByteDance/Seedream-4.5` | $0.042 | image |
| `google/flash-image-2.5` | $0.042 | image |
| `google/imagen-4.0-fast` | $0.042 | image |
| `google/imagen-4.0-ultra` | $0.126 | image |
| `hidream-ai/hidream-i1-fast` | $0.01575 | image |
| `hidream-ai/hidream-i1-full` | $0.042 | image |
| `ideogram/ideogram-3.0` | $0.063 | image |
| `openai/gpt-image-2` | from $0.0126 | image |
| `Qwen/Qwen-Image` | $0.0315 | image |
| `stabilityai/stable-diffusion-3-5-large` | $0.06825 | image |
| `Wan-AI/Wan2.6-Image-Edit` | $0.0315 | image |
| `Wan-AI/Wan2.6-T2I` | $0.0315 | image |

### Video generation

| Model | Price | Per |
| --- | --- | --- |
| `alibaba/happyhorse-1.0` | $0.294 | second |
| `bytedance/seedance-2.0-fast` | from $0.0735 | second |
| `google/veo-3.0-audio` | $0.42 | second |
| `google/veo-3.0-fast-audio` | $0.1575 | second |
| `kuaishou/kling-avatar-v2` | from $0.0588 | second |
| `kuaishou/kling-lip-sync` | $0.0147 | second |
| `kuaishou/kling-o1` | from $0.0882 | second |
| `kuaishou/kling-v2.5-turbo-pro` | $0.07875 | second |
| `kuaishou/kling-v2.6` | from $0.0735 | second |
| `kuaishou/kling-v3-motion-control` | from $0.0735 | second |
| `kuaishou/kling-v3-omni-video` | from $0.1764 | second |
| `kuaishou/kling-v3-video` | from $0.1764 | second |
| `minimax/hailuo-02` | from $0.105 | video |

### Text-to-speech

| Model | Price | Per |
| --- | --- | --- |
| `hexgrad/Kokoro-82M` | $0.0651 | 1K characters |
| `Qwen/Qwen3-TTS` | $0.105 | 1K characters |
| `ResembleAI/chatterbox-multilingual` | $0.105 | 1K characters |

### Transcription

| Model | Price | Per |
| --- | --- | --- |
| `mistralai/Voxtral-Small-24B-2507` | $0.00315 | minute |
| `openai/whisper-large-v3` | $0.04725 | minute |
| `openai/whisper-large-v3-turbo` | $0.021 | minute |

### Embeddings

| Model | Price | Per |
| --- | --- | --- |
| `BAAI/bge-large-en-v1.5` | $0.0105 | 1M tokens |
| `BAAI/bge-m3` | $0.0105 | 1M tokens |
| `google/embeddinggemma-300m` | $0.0021 | 1M tokens |
| `intfloat/multilingual-e5-large-instruct` | $0.0105 | 1M tokens |
| `nvidia/llama-nemotron-embed-vl-1b-v2` | $0.0105 | 1M tokens |
| `nvidia/Nemotron-3.5-Embed-8B-BF16` | $0.03675 | 1M tokens |
| `Qwen/Qwen3-Embedding-0.6B` | $0.0105 | 1M tokens |
| `Qwen/Qwen3-Embedding-4B` | $0.021 | 1M tokens |
| `Qwen/Qwen3-Embedding-8B` | $0.0105 | 1M tokens |