AI models
DeepSeek V4.1 Flash vs Gemini 3.8 Flash
Same criteria, side by side. Where the gap is small, it may be measurement noise.
Also searched as
DeepSeek V4 vs Gemini Flash
Direct answer: GPQA Diamond: DeepSeek V4.1 Flash 93.2%, Gemini 3.8 Flash 95.4% (Epoch AI). OTIS Mock AIME: DeepSeek V4.1 Flash 99.4%, Gemini 3.8 Flash 98.9% (Epoch AI). FrontierMath (tiers 1 to 3): DeepSeek V4.1 Flash 71.2%, Gemini 3.8 Flash 68.4% (Epoch AI). SimpleQA Verified: DeepSeek V4.1 Flash 42.9%, Gemini 3.8 Flash 69.7% (Epoch AI). GPQA Diamond (OpenRouter): DeepSeek V4.1 Flash 92.8%, Gemini 3.8 Flash 95.6% (OpenRouter Benchmarks). τ-bench Verified (airline): DeepSeek V4.1 Flash 77.8%, Gemini 3.8 Flash 76.7% (OpenRouter Benchmarks).
| Metric | DeepSeek V4.1 Flash | Gemini 3.8 Flash | Source |
|---|---|---|---|
| Weights ?what it is | open | closed | no data |
| Input US$/1M ? | US$ 0.022 | US$ 0.82 | no data |
| Output US$/1M ? | US$ 0.40 | US$ 4.12 | no data |
| Context ? | 1,048,576 | 1,048,576 | no data |
| GPQA Diamond ? | 93.2% | 95.4% | Epoch AI |
| OTIS Mock AIME ? | 99.4% | 98.9% | Epoch AI |
| FrontierMath (tiers 1 to 3) ? | 71.2% | 68.4% | Epoch AI |
| SimpleQA Verified ? | 42.9% | 69.7% | Epoch AI |
| GPQA Diamond (OpenRouter) ? | 92.8% | 95.6% | OpenRouter Benchmarks |
| τ-bench Verified (airline) ? | 77.8% | 76.7% | OpenRouter Benchmarks |
Compare with
Sources
- Epoch AI, CC BY 4.0, latest version we checked.
- OpenRouter Benchmarks, CC BY 4.0, latest version we checked.
- OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.
Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.