DeepSeek API pricing per million tokens
On 25 September 2026, DeepSeek models on this list cost from $0.03 per million input tokens (DeepSeek V4 Flash 0731) to $0.80 (R1 Distill Llama 70B), and from $0.0963 per million output tokens (DeepSeek V4 Flash 0423) to $2.50 (R1).
| Model | Input | Output | Cached input | Batch in / out | Context | Your month | Listed |
|---|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flash | $0.075 | $0.30 | $0.0015 | $0.112 / $0.336 | 1.04858M | $9.00 | 2026-09-10 |
| DeepSeek V4 Flash Vision Exp | $0.22 | $0.66 | $0.007 | 1.04858M | $23.10 | 2026-08-21 | |
| DeepSeek V4 Pro 0813 | $0.3458 | $1.04 | $0.0115 | 1.04858M | $36.31 | 2026-08-12 | |
| DeepSeek V4 Flash 0731 | $0.03 | $0.32 | $0.016 | 1.31072M | $6.60 | 2026-07-31 | |
| DeepSeek V4 Pro 0423 | $0.7129 | $1.43 | $0.0594 | 1.04858M | $64.16 | 2026-04-24 | |
| DeepSeek V4 Flash 0423 | $0.0482 | $0.0963 | $0.0096 | 1.04858M | $4.33 | 2026-04-24 | |
| DeepSeek V3.2 retiring 2026-09-28 | $0.269 | $0.40 | $0.1345 | 164K | $22.14 | 2025-12-01 | |
| DeepSeek V3.2 Exp retiring 2026-09-28 | $0.27 | $0.41 | 164K | $22.35 | 2025-09-29 | ||
| DeepSeek V3.1 Terminus retiring 2026-09-28 | $0.27 | $1.00 | $0.135 | 164K | $31.20 | 2025-09-22 | |
| DeepSeek V3.1 | $0.25 | $0.95 | $0.13 | 164K | $29.25 | 2025-08-21 | |
| R1 0528 | $0.50 | $2.15 | $0.35 | 164K | $62.25 | 2025-05-28 | |
| DeepSeek V3 0324 | $0.25 | $1.00 | 164K | $30.00 | 2025-03-24 | ||
| R1 Distill Llama 70B retiring 2026-09-28 | $0.80 | $0.80 | 8K | $60.00 | 2025-01-23 | ||
| R1 | $0.70 | $2.50 | 64K | $79.50 | 2025-01-20 | ||
| DeepSeek V3 | $0.32 | $0.89 | 164K | $32.55 | 2024-12-26 |
Compare DeepSeek with every other provider
Common questions
How much does the DeepSeek API cost?
On 25 September 2026, DeepSeek models on this list cost from $0.03 per million input tokens (DeepSeek V4 Flash 0731) to $0.80 (R1 Distill Llama 70B), and from $0.0963 per million output tokens (DeepSeek V4 Flash 0423) to $2.50 (R1).
What is the cheapest DeepSeek model?
For a typical mix of three input tokens to every output token, the cheapest is DeepSeek V4 Flash 0423 at $0.0482 per million input tokens and $0.0963 per million output tokens.
What is the newest DeepSeek model on the list?
DeepSeek V4.1 Flash, listed on 10 September 2026, at $0.075 per million input tokens and $0.30 per million output tokens.
Which DeepSeek model has the largest context window?
DeepSeek V4 Flash 0731, with 1.31072M tokens of context.
Does DeepSeek offer batch pricing?
Yes, 1 of its models on this list show a batch price, with batch input at a median of 149 percent of the standard price. Batch suits work that can wait for its answer.
Choosing a model for a production agent
Price is one input. Quality on your own cases, latency and the cost of retries decide the rest. Tell me what you are building.