Skip to main content

DeepSeek V3 vs DeepSeek-V4-Flash (Preview)

SpecDeepSeek V3DeepSeek-V4-Flash (Preview)
Respan Index13.2 (#166)32.3 (#82)
Blended price per 1M$0.175$0.056
Context window128K1M
Input / output per 1M$0.14 / $0.28$0.028 / $0.14
Max output8K-
Inputtexttext
Released2024-12-262026-04-24

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryDeepSeek V3DeepSeek-V4-Flash (Preview)
Coding3.1 DeepSeek V313.5 DeepSeek-V4-Flash (Preview)
Agents5.6* DeepSeek V343.2 DeepSeek-V4-Flash (Preview)
Reasoning17.3 DeepSeek V331.8 DeepSeek-V4-Flash (Preview)
Knowledge & language7.8 DeepSeek V323.6 DeepSeek-V4-Flash (Preview)
Human preference39.4 DeepSeek V351.9 DeepSeek-V4-Flash (Preview)

* Estimated: no published scores in that category. How the Respan Index works

Benchmarks both report

DeepSeek-V4-Flash (Preview) leads on 5 of the 5 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkDeepSeek V3DeepSeek-V4-Flash (Preview)
LMArena Elo13581439
GPQA Diamond59.1%88.1%
SimpleBench (SimpleBench)18.9%46.3%
MMLU-Pro75.9%86.2%
SWE-bench Verified42%79%

FAQ

Is DeepSeek V3 better than DeepSeek-V4-Flash (Preview)?

DeepSeek-V4-Flash (Preview) scores 32.3 on the Respan Index (#82), ahead of DeepSeek V3 at 13.2 (#166). DeepSeek-V4-Flash (Preview) leads on 5 of the 5 benchmarks both report.

Which is cheaper, DeepSeek V3 or DeepSeek-V4-Flash (Preview)?

DeepSeek V3 lists at $0.14 input and $0.28 output per 1M tokens; DeepSeek-V4-Flash (Preview) at $0.028 and $0.14. DeepSeek-V4-Flash (Preview) is 3.1x cheaper per token at list price.

Which has the larger context window?

DeepSeek V3 accepts 128K tokens and DeepSeek-V4-Flash (Preview) 1M tokens.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.