Eurus 70B
by OpenBMB / Tsinghua University · open-source · Last verified 2026-03-17
Eurus 70B is a suite of open-source reasoning-specialized models fine-tuned from LLaMA-2 70B using a comprehensive reward modeling approach covering helpfulness, correctness, coherence, complexity, and verbosity. It achieves state-of-the-art results among open-source models on a comprehensive benchmark suite spanning math, coding, and logical reasoning.
https://huggingface.co/openbmb/Eurus-70b-nca ↗C
C—Below Average
Adoption: CQuality: AFreshness: BCitations: BEngagement: F
Specifications
- License
- MIT
- Pricing
- open-source
- Capabilities
- mathematical-reasoning, code-reasoning, logical-reasoning, instruction-following
- Integrations
- Hugging Face, vLLM
- Use Cases
- complex math problem solving, competitive programming, logical reasoning chains, STEM education support
- API Available
- Yes
- Parameters
- ~70B
- Context Window
- 4K
- Modalities
- text
- Training Cutoff
- 2023
- Tags
- reasoning, llama, open-source, rlhf, math
- Added
- 2026-03-17
- Completeness
- 100%
Index Score
47.8Adoption
40
Quality
84
Freshness
62
Citations
60
Engagement
0