Replicate
- Categories
- GPU clouds, Inference hosts
- HQ
- US
- Website
- https://replicate.com
- Pricing page
- https://replicate.com/pricing
- Fee model
- per_token+per_second_gpu
- Fees
- Official LLMs billed per input/output token; most public models billed per second of hardware: CPU $0.000100/s, T4 $0.000225/s, L40S $0.000975/s, A100 80GB $0.001400/s, H100 $0.001525/s, H200 $0.001525/s (multi-GPU multiples). Private models billed for all online time incl. idle.
- OpenAI-compatible API
- no
- Notes
- Replicate predictions API (not OpenAI chat/completions shaped). Also resells proprietary LLMs (OpenAI, Anthropic, Google) per token.
- Sources
- https://replicate.com/pricing https://replicate.com/collections/language-models collected 2026-09-18
LLM prices (13 models, USD per 1M tokens)
| Model | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| DeepSeek R1 deepseek-ai/deepseek-r1 |
$3.75 | $10 | — | — | ✓ src |
| DeepSeek V3 deepseek-ai/deepseek-v3 |
$1.45 | $1.45 | — | — | ✓ src |
| DeepSeek V3.1 deepseek-ai/deepseek-v3.1 |
$0.672 | $2.02 | — | — | ✓ src |
| gpt-oss-120b openai/gpt-oss-120b |
$0.18 | $0.72 | — | — | ✓ src |
| gpt-oss-20b openai/gpt-oss-20b |
$0.09 | $0.36 | — | — | ✓ src |
| Kimi K2 Thinking moonshotai/kimi-k2-thinking |
$0.66 | $2.70 | — | — | ✓ src |
| Kimi K2.5 moonshotai/kimi-k2.5 |
$0.72 | $3.60 | — | — | ✓ src |
| Kimi K2.6 moonshotai/kimi-k2.6 |
$0.95 | $4 | — | — | ✓ src |
| Llama 3 70B Instruct meta/meta-llama-3-70b-instruct |
$0.65 | $2.75 | — | — | ✓ src |
| Llama 3 8B Instruct Lite meta/meta-llama-3-8b-instruct |
$0.05 | $0.25 | — | — | ✓ src |
| Llama 4 Maverick 17B Instruct meta/llama-4-maverick-instruct |
$0.25 | $0.95 | — | — | ✓ src |
| Llama 4 Scout 17B Instruct meta/llama-4-scout-instruct |
$0.17 | $0.65 | — | — | ✓ src |
| Qwen3 235B A22B 2507 qwen/qwen3-235b-a22b-instruct-2507 |
$0.264 | $1.06 | — | — | ✓ src |
GPU rates
| Provider | GPU | GPUs | VRAM/GPU | Billing | Per GPU-hr | Instance /hr | Region | Source |
|---|---|---|---|---|---|---|---|---|
| Replicate | T4 | 1 | 16 GB | serverless per sec | $0.81 | $0.81 | ✓ src ⓘ | |
| Replicate | L40S | 1 | 48 GB | serverless per sec | $3.51 | $3.51 | ✓ src ⓘ | |
| Replicate | L40S | 8 | 48 GB | serverless per sec | $3.51 | $28.08 | ✓ src ⓘ | |
| Replicate | A100 80GB | 1 | 80 GB | serverless per sec | $5.04 | $5.04 | ✓ src ⓘ | |
| Replicate | A100 80GB | 8 | 80 GB | serverless per sec | $5.04 | $40.32 | ✓ src ⓘ | |
| Replicate | H100 | 8 | 80 GB | serverless per sec | $5.49 | $43.92 | ✓ src ⓘ | |
| Replicate | H200 | 8 | — | serverless per sec | $5.49 | $43.92 | ✓ src ⓘ | |
| Replicate | H200 | 1 | — | serverless per sec | $5.49 | $5.49 | ✓ src ⓘ | |
| Replicate | H100 | 1 | 80 GB | serverless per sec | $5.49 | $5.49 | ✓ src ⓘ |