Model pricing · DeepSeek
DeepSeek V4 Pro pricing
DeepSeek V4 Pro costs $0.435 per million input tokens and $0.87 per million output tokens at list price. Best for: frontier value, 1m context.
| Input | $0.435 / 1M tokens |
|---|---|
| Output | $0.87 / 1M tokens |
| Cached input (read) | $0.003625 / 1M tokens |
| Context window | 1,000,000 tokens |
Cache-hit price as printed on provider page. Max output 384K.
Source: DeepSeek pricing, cross-checked against models.dev on 2026-09-15. Your provider invoice is authoritative.
What it costs in practice
| Workload | List price | If input is cached |
|---|---|---|
| One request: 10,000 tokens in, 1,000 out | $0.0052 | $0.0009 |
| A month of that request, 100 times a day | $15.66 | $2.72 |
| A 3,000-token instruction file on 100 requests a day, for a month | $3.92 | $0.033 |
The last row is the always-on cost of an instruction file like CLAUDE.md: paid on every request, whether the task needs it or not. Check how much of yours is waste.
Price history
From tokencheat's dated pricing snapshots. Only changes are listed.
| Since | Input / 1M | Output / 1M |
|---|---|---|
| 2026-01-01 | $0.435 | $0.87 |
Similarly priced models
Qwen3.7-Plus
$0.4 in · $1.6 out · Alibaba
Devstral 2
$0.4 in · $2 out · Mistral
Mistral Large 3
$0.5 in · $1.5 out · Mistral
Gemini 2.5 Flash
$0.3 in · $2.5 out · Google
Compare every model side by side or price your own workload.