DeepSeek R1 0528
Specifications
- Input
- Output
- Context window
- 164K tokens
- Released
- May 2025
Performance
- Speed
- 9 t/s
- TTFT
- 923 ms
- Latency
- —
- Intelligence
- —
Pricing
- Input
- €0.76 per 1M tokens
- Output
- €2.99 per 1M tokens
About this model
DeepSeek R1 0528 is an upgraded 685B parameter reasoning model with significantly enhanced depth of reasoning and inference capabilities. Achieves 87.5% on AIME 2025 (up from 70%), 91.4% on AIME 2024, 73.3% on LiveCodeBench, and 1930 Codeforces rating. Features system prompt support, averages 23K thinking tokens per question for deeper analysis, and reduced hallucination rate. Released under MIT license supporting commercial use and distillation. Performance approaching O3 and Gemini 2.5 Pro levels.
Technical specifications
- Capabilities
- Input modalities
- Output modalities
- Reasoning
- Default on
Knowledge horizon
Knowledge cutoff Jul 2024
Released May 2025
Today
Training to release 10 mo Since release 12 mo
See also
Add Model to Comparison
Search for a model to add