Transparency signal for adopters, based on public metrics:
• Adoption: 11.8k GitHub stars (public metric)
• Maintenance: last push ~1 month ago (active)
• Openness: weights + code fully public
• Docs: official README, on-device deploy guides
Not a rating of the science.
⚠️ Red Flags & Trade-offs (AI Verified)
- On-device small models still trail large models on complex multi-step reasoning, long-horizon tasks and broad world knowledge — expected trade-off. [GitHub]
- Multilingual OCR / translation is weaker (handwriting recognition is serial, and multi-language translation often falls back to English in practice). [GitHub]
- The Thinking variant trades latency for depth, losing the light, snappy response advantage; most performance numbers are reported by the authors (OpenBMB), with limited independent third-party benchmarking. [GitHub]
Each flag links to its primary source. See our Correction Policy.
📝 Report a correction or complaint
Corrections and complaints go straight to our editorial queue — read and actioned by the site operator, not lost in an inbox. No email needed.
Tsinghua ModelBest's on-device LLMs — SOTA small models engineered to run efficiently on phones and laptops.
11.8k stars · 818 forks · last push Sep 2026
💡 Lighter Alternatives (If this is too heavy)
Disclaimer
Red Flags and the AI score are editorial analysis, not endorsements. They are produced with AI assistance and reviewed by a human, but are not a substitute for reading the primary source. Scores reflect public adoption and availability signals only — never a verdict on the science, the code quality, or the people behind a project. We are not affiliated with any lab, university, or company named on this page, and links point only at primary sources (papers, repositories, journals). Content is informational and not engineering, business, or investment advice.