Maxwell V1 Flash FP8 is an open-source language model by maxwell-model. Features: 36.8b LLM, VRAM: 38.6GB, Context: 512K, License: apache-2.0.
| LLM Name | Maxwell V1 Flash FP8 |
| Repository π€ | https://huggingface.co/maxwell-model/Maxwell-V1-Flash-FP8 |
| Base Model(s) | |
| Model Size | 36.8b |
| Required VRAM | 38.6 GB |
| Updated | 2026-11-20 |
| Maintainer | maxwell-model |
| Model Type | maxwell_moe |
| Model Files | |
| Model Architecture | MaxwellMoeForCausalLM |
| License | apache-2.0 |
| Context Length | 524288 |
| Model Max Length | 524288 |
| Transformers Version | 5.16.0 |
| Tokenizer Class | PreTrainedTokenizerFast |
| Padding Token | <pad> |
| Vocabulary Size | 200704 |
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Maxwell V1 Flash | 512K / 0.6 GB | 287 | 144 |
| Maxwell V1 Flash Heretic | 512K / 0.6 GB | 142 | 0 |