See key specs and per-benchmark scores for each model/mode. Scroll horizontally for all columns. 当前对比 2 个模型的评测数据与核心参数。
DeepSeek-V4-Flash
DeepSeek-AI
DeepSeek-V4-Flash-Vision-Exp
DeepSeek-AI
Each axis is a category average, normalized to a 100-point radar.
Relative edge: none clear / Relative gap: 编程与软件工程 -4.2
Relative edge: 编程与软件工程 +4.2 / Relative gap: none clear
Method: for each model and benchmark, the chart first averages all scores in the current mode scope instead of taking the best score, then averages those benchmark scores within each category. Only benchmarks with at least two selected models scored are included; missing values are not counted as zero.
Best overall
DeepSeek-V4-Flash-Vision-Exp · 50.78
Best single
DeepSeek-V4-Flash-Vision-Exp · Terminal-Bench 2.1 83.90
Modality coverage
DeepSeek-V4-Flash-Vision-Exp · 2 modalities
Head to head
5
Benchmarks
0
Wins
5
Losses
-2.46
Average diff
Compare benchmark results across thinking modes and tool usage.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Complete scores for each model/mode across selected benchmarks.
5 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | DeepSeek-V4-Flash | DeepSeek-V4-Flash-Vision-Exp |
|---|---|---|
DeepSWE 编程与软件工程 | 54.40Thinking Level · High ｜ Tools | 59.30Thinking Level · High ｜ Tools |
NL2Repo-Bench 编程与软件工程 | 54.20Thinking Level · High ｜ Tools | 57.70Thinking Level · High ｜ Tools |
AutomationBench AI Agent - 工具使用 | 25.10Thinking Level · High ｜ Tools | 25.70Thinking Level · High ｜ Tools |
Terminal-Bench 2.1 AI Agent - 工具使用 | 82.70Thinking Level · High ｜ Tools | 83.90Thinking Level · High ｜ Tools |
Agents' Last Exam Agent能力评测 | 25.20Thinking Level · High ｜ Tools | 27.30Thinking Level · High ｜ Tools |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | DeepSeek-V4-FlashDeepSeek-AI | DeepSeek-V4-Flash-Vision-ExpDeepSeek-AI |
|---|---|---|
Core specsRelease | 2026-04-24 | 2026-08-21 |
Context length | 1M | 1M |
Parameters | 284B | — |
Active parameters | 13B | Not provided |
Max output | 384000 | 384000 |
MoE | Yes | No |
LicenseCode Open Source | Open Source · MIT License | Not provided |
Weights Open Source | Open Source · MIT License | Not provided |
Commercial use | 免费商用授权 | Not provided |
Local deploymentWeights | Hugging Face | Not provided |
Modality supportText Input/Output | / | / |
Image Input/Output | Not provided | / |
ResourcesPaper / report | DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence | DeepSeek-V4-Flash-Vision-Exp Release |
