DE
DeepSeek
AI model · China ·Apr 2026

DeepSeek V4-Flash

Open Source
284B
Parameters
1M
Context
5
Benchmarks

About DeepSeek V4-Flash

DeepSeek V4-Flash is an AI model developed by DeepSeek, released in 2026, made available as an open-source model with 284B params. DeepSeek V4-Flash is the smaller, latency-optimised member of the DeepSeek V4 family — 284B total parameters with 13B ac.

On this page you'll find DeepSeek V4-Flash's full specifications, including a context window of 1M tokens, 5 benchmark results. Review provider pricing and benchmark scores below, or compare DeepSeek V4-Flash head-to-head with other AI models.

→ More models by DeepSeek → AI benchmarks leaderboard

Links

Announcement ↗ Paper ↗ Hugging Face ↗

Community

Would you run DeepSeek V4-Flash again?
No votes yet — be the first.
Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

DeepSeek V4-Flash benchmark scores▶
Compare on the benchmark table →
Benchmark Category Score Variant
HumanEval code 91.3% pass@1
SWE-bench Verified code 72.1% thinking
MMLU-Pro general 78.9%
AIME 2025 math 85.4% thinking
GPQA Diamond reasoning 78.6% thinking

More models like DeepSeek V4-Flash

DE
DeepSeek V4-Pro
DeepSeek · 1.6T
 DE
DeepSeek-V3.2
DeepSeek
 DE
DeepSeek-V3.2-Speciale
DeepSeek
 DE
Deepseek OCR
DeepSeek · 3B
 DE
DeepSeek-V3.2-Exp
DeepSeek · 671B
 DE
DeepSeek-V3.1-Terminus
DeepSeek · 671B

Frequently asked questions

What is DeepSeek V4-Flash? ▶

DeepSeek V4-Flash is the smaller, latency-optimised member of the DeepSeek V4 family — 284B total parameters with 13B ac It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created DeepSeek V4-Flash? ▶

DeepSeek V4-Flash was developed by DeepSeek and released in 2026.

Is DeepSeek V4-Flash open source or proprietary? ▶

DeepSeek V4-Flash is released as open source — its weights are publicly available to download, run locally and fine-tune.

What is DeepSeek V4-Flash's context window? ▶

DeepSeek V4-Flash supports a context window of 1M tokens, and has 284B params.

How much does DeepSeek V4-Flash cost? ▶

Pricing for DeepSeek V4-Flash depends on the provider. See the providers table on this page for the latest API rates.

How does DeepSeek V4-Flash perform on benchmarks? ▶

DeepSeek V4-Flash is benchmarked across 5 evaluations on GenAIList, including HumanEval (91.3%). See the full benchmark table below and compare it with other models.

DeepSeek v4 in 4 Minutes
Deepseek v4: Best Opensource Model Ever? (Fully Tested)
DeepSeek V4 Review: Benchmarks, Pricing, And The 1.6T Parameter Whale
[FR] What you don't yet know about DeepSeek V4: faster, more powerful, 100% Open Source
DeepSeek V4: The Chinese monster that EXPLODES Opus 4.7

Know a great video about DeepSeek V4-Flash?