DeepSeek V4 Flash
Efficient million-token context intelligence at 13B activated parameters
Performance benchmarks
Model | AIME 2025 | GPQA Diamond | HLE | LiveCodeBench | MATH500 | SWE-bench verified |
|---|---|---|---|---|---|---|
DeepSeek V4 Flash | 91.60% | 79.00% | Related open-source models | Competitor closed-source models | ||
 Claude Opus 4.6 | 90.5% | 34.2% | 78.7% | |||
 OpenAI o3 | 83.3% | 24.9% | 99.2% | 62.3% | ||
OpenAI o1 | 76.8% | 96.4% | 48.9% | |||
GPT-4o | 49.2% | 2.7% | 32.3% | 89.3% | 31.0% |
This model is coming soon to Together’s Serverless API.
Deploy this model on an on-demand Dedicated Endpoint or pick a supported alternative from the Model Library.
- TypeReasoningChatCodeLLM
- Main use casesReasoning
- FeaturesFunction CallingJSON Mode
- IntelligenceHigh
- Parameters284B
- Activated parameters13B
- Context length1M
- Quantization levelFP4
- External link
- CategoryChat