- Terminal Bench 2.1 83.9
- NL2Repo 57.7
- Cybergym 75.3
- DeepSWE 59.3
- Toolathlon-Verified 75.9
- DSBench-Hard (in-house) 63.6
- AutomationBench (Public) 25.7
- Agents Last Exam 27.3
- Chartography 64.3
- ZeroBench (Pass@5) 35
DeepSeek-V4-Flash-Vision-Exp
DeepSeek-V4-Flash-Vision-Exp places in the top three on Toolathlon.
11
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
DeepSeek Aug 2026 DeepSeek-V4-Flash-Vision-Exp
Open weights Mid-size
No Epoch Capabilities Index score yet
- NL2Repo-Bench 57.7% #7 /28
- Vision2Web 67.6% #8 /10
- Chartography 64.3% #8 /12
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| NL2Repo-Bench | 57.7% | #7 of 28 | Model card | |
| DeepSWE | 59.3% | #37 of 51 | Model card | |
| Vision2Web | 67.6% | #8 of 10 | Official leaderboard | |
| DSBench (DeepSeek internal) | 63.6% | #3 of 3 | Model card |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Toolathlon | 75.9% | #3 of 37 | Model card | |
| Terminal-Bench 2.1 | 83.9% | #18 of 53 | Model card | |
| AutomationBench | 25.7% | #32 of 41 | Model card |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Agents' Last Exam | 27.3% | #16 of 25 | Model card |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| ZeroBench | 35% | #7 of 25 | Model card | |
| Chartography | 64.3% | #8 of 12 | Model card |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| CyberGym | 75.3% | #21 of 33 | Model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.