Muse 2 350M is an open-source language model by Muse-research. Features: 350m LLM, VRAM: 1.4GB, Context: 32K, License: apache-2.0, LLM Explorer Score: 0.33.
En Eval-results Grouped-query-attention Long-context Muse-research Muse nova transformer Pytorch Reasoning Region:us Safetensors Science
Model Card on HF 🤗: https://huggingface.co/Muse-research/Muse-2-350M
Muse 2 350M Benchmarks
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Muse 2 350M Parameters and Internals
| LLM Name | Muse 2 350M |
| Repository 🤗 | https://huggingface.co/Muse-research/Muse-2-350M |
| Model Size | 350m |
| Required VRAM | 1.4 GB |
| Updated | 2026-06-08 |
| Maintainer | Muse-research |
| Model Type | muse_nova_transformer |
| Model Files | |
| Supported Languages | en |
| Model Architecture | MuseNovaForCausalLM |
| License | apache-2.0 |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Tokenizer Class | TokenizersBackend |
| Padding Token | <|pad|> |
| Vocabulary Size | 32768 |
Rank the Muse 2 350M Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 54413 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a