DeepSeek V4 Flash
Tool useReasoningPrompt cachingStructured output
Context
1M
Max output
384K
Serving providers
0
Cheapest input
n/a /1M
Benchmarks
independent evaluations · source: Artificial Analysis and Hugging Face leaderboardsSpeed & latency
Output speed
111 t/s
tokens / second
Time to first token
3.47s
latency
Headline indices
Individual benchmarks · click a row for effort variants
| Benchmark | Best score | Effort | Cost / task | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AA-Briefcase | 16% | — | — | |||||||||||||||||
| CritPt | 7% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| EvalComputeProxy | 4874.7 | default | — | |||||||||||||||||
| ||||||||||||||||||||
| GDPval | 1188.5 | default | — | |||||||||||||||||
| GPQA Diamond | 89% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Humanity's Last Exam | 32% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| IFBench | 79% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| IT-Bench SRE | 32% | default | — | |||||||||||||||||
| Long-Context Reasoning | 63% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| Omniscience | -22.3 | high | — | |||||||||||||||||
| ||||||||||||||||||||
| SciCode | 45% | default | — | |||||||||||||||||
| ||||||||||||||||||||
| SWE-bench Verified | 79% | default | — | |||||||||||||||||
| TerminalBench Hard | 39% | high | — | |||||||||||||||||
| ||||||||||||||||||||
| TerminalBench v2.1 | 62% | default | — | |||||||||||||||||
| τ-bench Banking | 23% | default | — | |||||||||||||||||
| τ²-bench | 96% | high | — | |||||||||||||||||
| ||||||||||||||||||||
Pricing by serving provider
availabilityNo public price yet. Not offered on-demand at any tracked provider.
Price history
input + output $/1M since we started tracking Input Output
Input down 29% since first tracked
Frequently asked questions
DeepSeek V4 Flash pricing, context and availabilityWhat is the context window of DeepSeek V4 Flash?
DeepSeek V4 Flash accepts up to 1M tokens of context and can return up to 384K output tokens.
What can DeepSeek V4 Flash do?
DeepSeek V4 Flash supports tool use, reasoning, prompt caching and structured output.
Other DeepSeek models
compare pricing across the DeepSeek lineup Pricing verified against LiteLLM + OpenRouter + provider pages · refreshed daily
Every weekday
AI moves fast. Here's your debrief.
News, analysis, tools, and more.
For people who build with AI