Sarv Reasoning is an open-source language model by artindnr. Features: 20.9b LLM, VRAM: 41.6GB, Context: 128K, License: apache-2.0, LLM Explorer Score: 0.3.
| LLM Name | Sarv Reasoning |
| Repository π€ | https://huggingface.co/artindnr/sarv-reasoning |
| Base Model(s) | |
| Model Size | 20.9b |
| Required VRAM | 41.6 GB |
| Updated | 2026-08-11 |
| Maintainer | artindnr |
| Model Type | gpt_oss |
| Model Files | |
| Supported Languages | fa |
| Model Architecture | GptOssForCausalLM |
| License | apache-2.0 |
| Context Length | 131072 |
| Model Max Length | 131072 |
| Transformers Version | 5.14.1 |
| Tokenizer Class | TokenizersBackend |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 201088 |
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| ChatBerry 1.1 | 128K / 41.6 GB | 265 | 2 |
| Sarv Hybrid Pro | 128K / 41.6 GB | 115 | 1 |
| ChatBerry | 128K / 41.6 GB | 116 | 3 |
| ChatBerry 1 | 128K / 41.6 GB | 161 | 2 |
| Sarv Hybrid | 128K / 41.6 GB | 119 | 1 |
| Sarv Non Reasoning | 128K / 41.6 GB | 118 | 1 |
| GPT Oss 20B Abliterated | 128K / 41.6 GB | 12262 | 4 |
| Context 1 MLX MXFP4 | 128K / 11.8 GB | 12 | 2 |
| ...ful Pruned MoE Only 32 Experts | 128K / 40.8 GB | 66 | 3 |
| Context 1 MLX 4bit | 128K / 11.8 GB | 30 | 1 |