DeepSeek V3 vs DeepSeek V4 Flash
| Spec | DeepSeek V3 | DeepSeek V4 Flash |
|---|---|---|
| Respan Index | 13.2 (#166) | 45.4 (#54) |
| Blended price per 1M | $0.175 | $0.113 |
| Context window | 128K | 1M |
| Input / output per 1M | $0.14 / $0.28 | $0.09 / $0.18 |
| Max output | 8K | - |
| Input | text | text |
| Released | 2024-12-26 | 2026-07-31 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | DeepSeek V3 | DeepSeek V4 Flash |
|---|---|---|
| Coding | 3.1 DeepSeek V3 | 29.7 DeepSeek V4 Flash |
| Agents | 5.6* DeepSeek V3 | 58.0 DeepSeek V4 Flash |
| Reasoning | 17.3 DeepSeek V3 | 46.1 DeepSeek V4 Flash |
| Knowledge & language | 7.8 DeepSeek V3 | 35.5 DeepSeek V4 Flash |
| Human preference | 39.4 DeepSeek V3 | 55.9 DeepSeek V4 Flash |
* Estimated: no published scores in that category. How the Respan Index works
Benchmarks both report
DeepSeek V4 Flash leads on 5 of the 5 benchmarks both report. Higher is better. Each score links to where it was published.
FAQ
Is DeepSeek V3 better than DeepSeek V4 Flash?
DeepSeek V4 Flash scores 45.4 on the Respan Index (#54), ahead of DeepSeek V3 at 13.2 (#166). DeepSeek V4 Flash leads on 5 of the 5 benchmarks both report.
Which is cheaper, DeepSeek V3 or DeepSeek V4 Flash?
DeepSeek V3 lists at $0.14 input and $0.28 output per 1M tokens; DeepSeek V4 Flash at $0.09 and $0.18. DeepSeek V4 Flash is 1.6x cheaper per token at list price.
Which has the larger context window?
DeepSeek V3 accepts 128K tokens and DeepSeek V4 Flash 1M tokens.
Compare other models
- vsDeepSeek V3.1 vs DeepSeek V3
- vsDeepSeek V4.1 Flash vs DeepSeek V4 Flash
- vsMuse Spark 1.3 vs DeepSeek V4 Flash
- vsGPT-6 Luna vs DeepSeek V4 Flash
- vsDeepSeek V3 vs gpt-oss-20b
- vsDeepSeek V3 vs Llama 4 Scout
- vsDeepSeek V3 vs Qwen3 30B-A3B Instruct 2507
- vsDeepSeek V4 Flash vs Gemini 3.6 Flash
- vsDeepSeek V4 Flash vs GLM-5.3-Flash
- vsDeepSeek V4 Flash vs Muse Spark 1.2