deepseek-v4-flash
DeepSeek V4 Flash · DeepSeek
- Multiplier
- ×0.31
- Context served
- not published
- Max output
- not published
DeepSeek V4 Flash from DeepSeek, served on an OpenAI-compatible endpoint. This model is retired and no longer accepted. The page stays up because the measurements on it are still true of the model.
The same record /models renders, read from the row that serves the next request for this slug.
DeepSeek V4 Flash · DeepSeek
Behaviour that is not in the provider's documentation because nobody wrote it down. Each page names the probe and the date.
On GLM, yes, measurably. On DeepSeek and Kimi the parameter is accepted and the output does not change.
Not while reasoning is on. The request is refused rather than downgraded.
All five report cache hits under usage.prompt_tokens_details.cached_tokens, and that is the field we bill from.
No. A POST to /v1/responses returns 404. Use /v1/chat/completions.
Because max_tokens budgets the reasoning too, and the reasoning is emitted first. Every model here returned an empty answer at 512 and below.
Three of the five see images. GLM-5.3 refuses the request outright. DeepSeek-V4-Pro accepts it, returns 200, and does not see the image.
Every model is billed in weighted tokens against one multiplier, so the arithmetic is the same whichever you pick. How a weighted token is computed walks through it, and pricing has the tiers.
Cached input is the lever worth knowing about on this model as much as any other. Cache hits are reported back to you in the response body, in the same field we bill from, and on a long agent context they are the difference between most of a request and a fraction of it. See prompt caching.