# DeepSeek R1 API Pricing

As of October 10, 2026, DeepSeek R1 costs $0.55 per 1M input tokens and $2.19 per 1M output tokens on DeepSeek, with cache reads at $0.14 per 1M tokens. Across 16 providers on LiteLLM, the input price ranges from $0.20 to $5.00. It has a 65,536-token context window and returns up to 8,192 tokens per response.

- Page: https://models.litellm.ai/models/deepseek-r1
- Lab: DeepSeek
- LiteLLM model name: `deepseek/deepseek-r1`
- Context window: 65,536 tokens
- Max output: 8,192 tokens
- Added to LiteLLM: January 2025
- Prices updated: October 10, 2026

## Pricing by provider

Token prices in USD per 1M tokens.

| Provider | Model name | Input | Output | Cache read | Per image | Per second | Context |
|---|---|---|---|---|---|---|---|
| DeepSeek | `deepseek/deepseek-r1` | $0.55 | $2.19 | $0.14 | – | – | 64K |
| Amazon Bedrock | `deepseek.r1-v1:0` | $1.35 | $5.40 | – | – | – | 128K |
| Azure AI Foundry | `azure_ai/deepseek-r1` | $1.35 | $5.40 | – | – | – | – |
| Crusoe (snapshot 0528) | `crusoe/deepseek-ai/DeepSeek-R1-0528` | $3.00 | $7.00 | – | – | – | 160K |
| DeepInfra | `deepinfra/deepseek-ai/DeepSeek-R1` | $0.70 | $2.40 | – | – | – | 160K |
| DeepInfra (snapshot 0528) | `deepinfra/deepseek-ai/DeepSeek-R1-0528` | $0.50 | $2.15 | $0.40 | – | – | 160K |
| Fireworks AI | `fireworks_ai/accounts/fireworks/models/deepseek-r1` | $3.00 | $8.00 | – | – | – | 128K |
| Fireworks AI (snapshot 0528) | `fireworks_ai/accounts/fireworks/models/deepseek-r1-0528` | $3.00 | $8.00 | – | – | – | 160K |
| Google Vertex AI (snapshot 0528) | `vertex_ai/deepseek-ai/deepseek-r1-0528-maas` | $1.35 | $5.40 | – | – | – | 65K |
| Hyperbolic | `hyperbolic/deepseek-ai/DeepSeek-R1` | $0.40 | $0.40 | – | – | – | 32K |
| Hyperbolic (snapshot 0528) | `hyperbolic/deepseek-ai/DeepSeek-R1-0528` | $0.25 | $0.25 | – | – | – | 128K |
| Lambda (snapshot 0528) | `lambda_ai/deepseek-r1-0528` | $0.20 | $0.60 | – | – | – | 128K |
| Nebius AI Studio | `nebius/deepseek-ai/DeepSeek-R1` | $0.80 | $2.40 | – | – | – | 128K |
| Nebius AI Studio (snapshot 0528) | `nebius/deepseek-ai/DeepSeek-R1-0528` | $0.80 | $2.40 | – | – | – | 164K |
| Novita AI | `novita/deepseek/deepseek-r1` | $4.00 | $4.00 | – | – | – | 64K |
| Novita AI (snapshot 0528) | `novita/deepseek/deepseek-r1-0528` | $0.70 | $2.50 | $0.35 | – | – | 160K |
| Replicate | `replicate/deepseek-ai/deepseek-r1` | $3.75 | $10.00 | – | – | – | 64K |
| SambaNova | `sambanova/DeepSeek-R1` | $5.00 | $7.00 | – | – | – | 32K |
| Snowflake Cortex | `snowflake/deepseek-r1` | $1.35 | $5.40 | – | – | – | 128K |
| Together AI (snapshot 0528) | `together_ai/deepseek-ai/DeepSeek-R1-0528` | $3.00 | $7.00 | – | – | – | 160K |
| Vercel AI Gateway | `vercel_ai_gateway/deepseek/deepseek-r1` | $0.55 | $2.19 | – | – | – | 128K |
| Amazon Bedrock (US cross-region) | `us.deepseek.r1-v1:0` | $1.35 | $5.40 | – | – | – | 128K |

## Cost examples

1,000 requests with 2,000 input and 500 output tokens each:

- DeepSeek: $2.195
- Lambda: $0.70
- Hyperbolic: $0.625
- DeepInfra: $2.075

## Features

Supported: Function calling, Parallel tool calls, Structured output, Reasoning, Prompt caching.

## Use it with LiteLLM

```python
from litellm import completion

response = completion(
    model="deepseek/deepseek-r1",
    messages=[{"role": "user", "content": "Hello!"}],
)
```

## FAQ

### How much does DeepSeek R1 cost?

DeepSeek R1 costs $0.55 per 1M input tokens and $2.19 per 1M output tokens on DeepSeek, with cache reads at $0.14 per 1M tokens. Across 16 providers on LiteLLM, the input price ranges from $0.20 to $5.00.

### What is the context window of DeepSeek R1?

DeepSeek R1 has a 65,536-token context window and returns up to 8,192 tokens per response.

### Which providers offer DeepSeek R1?

DeepSeek R1 is available from 16 providers through LiteLLM: DeepSeek, Amazon Bedrock, Azure AI Foundry, Crusoe, DeepInfra, Fireworks AI, Google Vertex AI, Hyperbolic, Lambda, Nebius AI Studio, Novita AI and Replicate and 4 more. The lowest price is on Lambda.

### How do I call DeepSeek R1 with an OpenAI-compatible API?

Use the LiteLLM Python SDK with model="deepseek/deepseek-r1", or add the model to the LiteLLM proxy's config.yaml and send requests to /v1/chat/completions from any OpenAI client. LiteLLM tracks the cost of every request at the prices on this page.

### What features does DeepSeek R1 support?

DeepSeek R1 supports function calling, parallel tool calls, structured output, reasoning and prompt caching.

### When did LiteLLM add DeepSeek R1?

DeepSeek R1 was added to LiteLLM's model price file in January 2025.

Source: LiteLLM model price file, https://github.com/BerriAI/litellm/blob/main/model_prices_and_context_window.json