Model comparison
DeepSeek-V4-Pro vs DeepSeek-V4-Flash
- Input
- $0.66 off-peak · $1.32 peak
- Output
- $1.98 off-peak · $3.96 peak
- Context
- 1M
- Released
- 2026-08-13
- Input
- $0.22 off-peak · $0.44 peak
- Output
- $0.66 off-peak · $1.32 peak
- Context
- 1M
- Released
- 2026-04-24
DeepSeek-V4-Pro scores higher on 6 of the 6 benchmarks both models report; DeepSeek-V4-Flash is cheaper on input tokens ($0.2 vs $0.7 per 1M, 3×).
Specs and pricing side by side
| Spec | DeepSeek-V4-Pro | DeepSeek-V4-Flash |
|---|---|---|
| Maker | DeepSeek | DeepSeeksame |
| Released | 2026-08-13 | 2026-04-24 |
| Context window | 1M | 1Msame |
| Max output | 384K | 384Ksame |
| Input price | $0.66 off-peak · $1.32 peak | $0.22 off-peak · $0.44 peak |
| Output price | $1.98 off-peak · $3.96 peak | $0.66 off-peak · $1.32 peak |
| Parameters | 1.6T total · 49B active | 284B total · 13B active |
| License | MIT | MITsame |
| Weights | Open weights | Open weightssame |
| Modalities | Text | Textsame |
Not published by either maker: knowledge cutoff.
Benchmarks both models report
| Benchmark | DeepSeek-V4-Pro | DeepSeek-V4-Flash |
|---|---|---|
| Terminal Bench 2.1 | 87.9% | 82.7% |
| Cybergym | 83.3% | 76.7% |
| Toolathlon-Verified | 74.1% | 70.3% |
| DSBench-FullStack | 71.1% | 68.7% |
| DeepSWE | 62.7% | 54.4% |
| NL2Repo | 61.5% | 54.2% |
Scores are as published by each maker or its cited source; harness, reasoning effort and date can differ between rows, so treat small gaps as ties. Sources are listed on each model page.
Which one should you pick?
Pick DeepSeek-V4-Pro if…
- You care about Terminal Bench 2.1, Cybergym, Toolathlon-Verified (it scores higher on 6 shared benchmarks).
- Your use case is: long-document and large-codebase analysis that needs the full 1M-token window.
Pick DeepSeek-V4-Flash if…
- Cost matters: input tokens cost 3× less ($0.2 vs $0.7 per 1M).
- Your use case is: high-volume agentic and coding workflows where cost and latency matter more than absolute peak quality.
FAQ
Which is better, DeepSeek-V4-Pro or DeepSeek-V4-Flash?
DeepSeek-V4-Pro scores higher on 6 of the 6 benchmarks both models report; DeepSeek-V4-Flash is cheaper on input tokens ($0.2 vs $0.7 per 1M, 3×). The better pick depends on whether you value benchmark scores, price per token, context length or openness for your workload.
How much cheaper is DeepSeek-V4-Flash?
DeepSeek-V4-Pro: $0.66 off-peak · $1.32 peak input, $1.98 off-peak · $3.96 peak output. DeepSeek-V4-Flash: $0.22 off-peak · $0.44 peak input, $0.66 off-peak · $1.32 peak output (list API prices as published by each maker).
What are the context windows of DeepSeek-V4-Pro and DeepSeek-V4-Flash?
DeepSeek-V4-Pro offers 1M with up to 384K output; DeepSeek-V4-Flash offers 1M with up to 384K output.
Is DeepSeek-V4-Pro newer than DeepSeek-V4-Flash?
DeepSeek-V4-Pro was released 2026-08-13; DeepSeek-V4-Flash was released 2026-04-24.