DeepSeek-V4-Flash-Vision-ExpvsDeepSeek-V4-Flash
Across 5 shared benchmarks, DeepSeek-V4-Flash-Vision-Exp leads overall: DeepSeek-V4-Flash-Vision-Exp wins 5, DeepSeek-V4-Flash wins 0, with 0 ties and an average score difference of +2.46.
DeepSeek-V4-Flash-Vision-Exp
DeepSeek-AI · 2026-08-21 · Multimodal model
DeepSeek-V4-Flash
DeepSeek-AI · 2026-04-24 · Reasoning model
DeepSeek-V4-Flash-Vision-Exp5 wins(100%)(0%)0 winsDeepSeek-V4-Flash
Benchmark scores
Grouped by capability, sorted by largest gap within each. 5 shared benchmarks.
AI Agent - Tool Usage
DeepSeek-V4-Flash-Vision-Exp 2/2| Benchmark | DeepSeek-V4-Flash-Vision-Exp | DeepSeek-V4-Flash | Diff |
|---|---|---|---|
| Terminal-Bench 2.1 | 83.9011 / 44Max (With Tools) | 82.7016 / 44Max (With Tools) | +1.20 |
| AutomationBench | 25.707 / 8Max (With Tools) | 25.108 / 8Max (With Tools) | +0.60 |
Coding and Software Engineer
DeepSeek-V4-Flash-Vision-Exp 2/2| Benchmark | DeepSeek-V4-Flash-Vision-Exp | DeepSeek-V4-Flash | Diff |
|---|---|---|---|
| DeepSWE | 59.3012 / 27Max (With Tools) | 54.4015 / 27Max (With Tools) | +4.90 |
| NL2Repo-Bench | 57.703 / 8Max (With Tools) | 54.205 / 8Max (With Tools) | +3.50 |
Agent Level Benchmark
DeepSeek-V4-Flash-Vision-Exp 1/1| Benchmark | DeepSeek-V4-Flash-Vision-Exp | DeepSeek-V4-Flash | Diff |
|---|---|---|---|
| Agents' Last Exam | 27.306 / 11Max (With Tools) | 25.2010 / 11Max (With Tools) | +2.10 |
Specs
| Field | DeepSeek-V4-Flash-Vision-Exp | DeepSeek-V4-Flash |
|---|---|---|
| Publisher | DeepSeek-AI | DeepSeek-AI |
| Release date | 2026-08-21 | 2026-04-24 |
| Model type | Multimodal model | Reasoning model |
| Architecture | Dense | MoE |
| Parameters | Not available | 284B |
| Context length | 1M | 1M |
| Max output | 384K | 384K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | DeepSeek-V4-Flash-Vision-Exp | DeepSeek-V4-Flash |
|---|---|---|
| Text input | $0.22 / 1M tokens | $0.14 / 1M tokens |
| Text output | $0.66 / 1M tokens | $0.28 / 1M tokens |
| Cache read | $0.007 / 1M tokens | $0.0028 / 1M tokens |
Summary
- DeepSeek-V4-Flash-Vision-Expleads in:AI Agent - Tool Usage (2/2), Coding and Software Engineer (2/2), Agent Level Benchmark (1/1)
On average across the 5 shared benchmarks, DeepSeek-V4-Flash-Vision-Exp scores 2.46 higher.
Largest single-benchmark gap: DeepSWE — DeepSeek-V4-Flash-Vision-Exp 59.30 vs DeepSeek-V4-Flash 54.40 (+4.90).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.