All models / DeepSeek
DeepSeek R1
Reasoning model with chain-of-thought output. MIT-licensed open weights.
At a glance
- Provider
- DeepSeek
- Family
- deepseek-r1
- Released
- 2025-01
- Status
- Active
- Context window
- 128K
- Max output
- 8K
- Knowledge cutoff
- 2024-07
- API id
- deepseek-reasoner
Pricing
USD per 1M tokens.
- Input
- $0.550
- Output
- $2.19
- Cache read
- $0.140
- Cache write (5m)
- —
- Cache write (1h)
- —
- Batch input
- —
- Batch output
- —
Capabilities
- Tool use
- No
- Structured outputs
- No
- Vision
- No
- Extended thinking
- Yes
- Prompt caching
- Yes
- Batch API
- No
- Fine-tuning
- No
- Open weights
- Yes
Modalities
- Input
- text
- Output
- text
Compare with
Links
Last verified 2026-04-26. License: mit.