Trinity Nano Preview MLX 8bit is an open-source language model by arcee-ai. Features: 6.1b LLM, VRAM: 6.4GB, Context: 128K, License: other, Quantized, LLM Explorer Score: 0.22.
| LLM Name | Trinity Nano Preview MLX 8bit |
| Repository 🤗 | https://huggingface.co/arcee-ai/Trinity-Nano-Preview-MLX-8bit |
| Base Model(s) | |
| Model Size | 6.1b |
| Required VRAM | 6.4 GB |
| Updated | 2026-08-09 |
| Maintainer | arcee-ai |
| Model Type | afmoe |
| Model Files | |
| Supported Languages | en es fr de it pt ru ar hi ko zh |
| Quantization Type | 8bit |
| Model Architecture | AfmoeForCausalLM |
| License | other |
| Context Length | 131072 |
| Model Max Length | 131072 |
| Transformers Version | 4.57.3 |
| Tokenizer Class | TokenizersBackend |
| Padding Token | <|pad|> |
| Vocabulary Size | 200192 |
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Trinity Nano Preview | 128K / 12.2 GB | 18452 | 78 |
| Dolphin X1 Trinity Nano | 128K / 12.2 GB | 151 | 30 |
| Trinity Nano Base | 128K / 12.2 GB | 1891 | 29 |
| Trinity Nano Preview FP8 Block | 128K / 6.7 GB | 325 | 1 |
| Trinity Nano Base Pre Anneal | 4K / 12.2 GB | 66 | 14 |