Nanbeige
Nanbeige ranks #55 of 91 providers on BenchLeader by its best model, with 1 ranked model. Its best model is Nanbeige4.1-3B at 40.9 on the BenchLeader Index, strongest in knowledge. 1 of its 3 models ship open weights.
- Provider rank
- #55
- by best model
- Best index
- 40.9
- Nanbeige4.1-3B
- Ranked models
- 1
- 1 open weights
- Cheapest ranked
- –
Release timeline
Each model’s best configuration by release date. The top eight are labelled; hover any point for its name.
Not enough dated, ranked models to draw a timeline.
Next release
Not enough dated releases to estimate a cadence.
Models
Every Nanbeige model with a result, one row per model; switch to all configurations to see each reasoning effort.
3 of 3
| # | ||||||||
|---|---|---|---|---|---|---|---|---|
| 1 | Nanbeige4.1-3Bopen | 40.9±5 | – | – | – | 256k | 2026-02-11 | 2026-02-11 · stale |
| 2 | Nanbeige3.5 Prothinking | – | – | – | – | – | – | – |
| 3 | Nanbeige4 3Bthinking | – | – | – | – | – | – | – |
Latest news
- OpenAI Astra “Secret Technique” Actually a Decade Oldflyingpenguin · 3 Sept 2026
- Inside Nanbeige4.2-3B-Base’s Looped Transformer ArchitectureHackerNoon · 28 Jul 2026
- Nanbeige4.2-3B vLLM support is landing — fork-only era endsOrcaRouter · 9 Sept 2026
- The 'LLM Architecture Gallery' illustrates the architectures of various large-scale language models such as GPT, Llama, and Grok.GIGAZINE · 17 Mar 2026
- Nanbeige4-3B-Thinking: How a 23T Token Pipeline Pushes 3B Models Past 30B Class ReasoningMarkTechPost · 12 Dec 2025