The open-weights model achieves competitive results on AIME and MATH benchmarks using reinforcement learning without initial supervised fine-tuning.
The open-weights model achieves competitive results on AIME and MATH benchmarks using reinforcement learning without initial supervised fine-tuning.
VentureBeat
Original Reporting Source
Explore features, ratings, and alternatives from our 19,000+ AI tools catalog.

Frontier open-weights reasoning model matching OpenAI o1 with DeepSeek-R1 and DeepSeek-V3 MoE architecture.