At a glance
The older names deepseek-v4-flash and deepseek-v4-flash-vision-exp are still accepted; DeepSeek serves them with V4.1 Flash and bills them at the Flash price.
| DeepSeek V4.1 Flash | |
|---|---|
| API model name | deepseek-flash |
| Model version | DeepSeek-V4.1-Flash-1017 |
| Context length | 1M tokens |
| Maximum output | Up to 384K tokens |
| Thinking | Thinking and non-thinking modes; thinking is the default |
| Tool calls, JSON output, Anthropic-compatible API | Supported |
| Vision | Supported |
| Concurrency limit | 2,500 |
API pricing
Prices per million tokens on DeepSeek's API. Off-peak rates are half the peak rates. Peak hours are 01:00 to 04:00 and 06:00 to 10:00 UTC, Monday to Friday, excluding Chinese public holidays; every other hour is off-peak.
| Tokens | Peak | Off-peak | V4 Pro, peak |
|---|---|---|---|
| Input, cache hit | $0.006 | $0.003 | $0.044 |
| Input, cache miss | $0.30 | $0.15 | $1.32 |
| Output | $1.20 | $0.60 | $3.96 |
What a coding session costs
Coding agents resend the conversation on every turn, so most input is a cache hit. Take an illustrative turn with 100,000 input tokens, 90% of them cached, and 2,000 output tokens, at peak prices:
- 90,000 cached input tokens at $0.006 per million: $0.00054
- 10,000 uncached input tokens at $0.30 per million: $0.003
- 2,000 output tokens at $1.20 per million: $0.0024
- About $0.006 per turn, so roughly $2.97 for 500 turns at peak, or about $1.49 off-peak
Use it in Claude Code
DeepSeek's API has an Anthropic-compatible endpoint at https://api.deepseek.com/anthropic; use deepseek-flash as the model name. Nexum Router serves the same model on its Claude Code endpoint as deepseek-v4.1-flash. Both setups, with the exact settings, are in How to use DeepSeek in Claude Code.
Use it in opencode
- opencode Go includes DeepSeek V4.1 Flash with an $80 monthly allowance, capped at 20% of it per five hours and 50% per week (comparison)
- Any OpenAI-compatible provider works too, including DeepSeek's API and Nexum Router: DeepSeek in opencode
DeepSeek V4.1 Flash on Nexum Router
Every Nexum Router plan includes $5 of DeepSeek V4.1 Flash usage a day, renewed daily, with top-ups from the dashboard. The model ID is deepseek-v4.1-flash on both the OpenAI-compatible and the Claude Code endpoint.
Frequently asked questions
What is DeepSeek V4.1 Flash called in the API?
deepseek-flash. The older name deepseek-v4-flash is still accepted and served by V4.1 Flash.
How much does DeepSeek V4.1 Flash cost?
$0.30 per million input tokens on a cache miss, $0.006 on a cache hit, and $1.20 per million output tokens at peak. Off-peak prices are half.
What is DeepSeek V4.1 Flash's context window?
1M tokens on DeepSeek's API, with up to 384K tokens of output.
Does DeepSeek V4.1 Flash work with Claude Code?
Yes, through DeepSeek's Anthropic-compatible endpoint or through another Anthropic-compatible provider such as Nexum Router.
Sources
Prices, limits and features were checked against these pages on 2026-10-29.