TinyMistral 248M V3 is an open-source language model by M4-ai. Features: 248m LLM, VRAM: 0.5GB, Context: 32K, License: apache-2.0, LLM Explorer Score: 0.13.
Conversational Dataset:eminorhan/gutenberg en Dataset:jeankaddour/minipile Dataset:llm360/txt360 Dataset:locutusque/tm-data-v2 Dataset:mlfoundations/dclm-bas... Dataset:skylion007/openwebtext En Endpoints compatible Mistral Model-index Region:us Safetensors
Model Card on HF 🤗: https://huggingface.co/M4-ai/TinyMistral-248M-v3
TinyMistral 248M V3 Benchmarks
MMLU Pro: 1.47
MUSR: 5.15
BBH: 1.78
IFEval: 16.39 vs 88 (so35)-81.4%
MATH Lvl 5: 0.45
LLME Score: 0.12894
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)
TinyMistral 248M V3 Parameters and Internals
| LLM Name | TinyMistral 248M V3 |
| Repository 🤗 | https://huggingface.co/M4-ai/TinyMistral-248M-v3 |
| Model Size | 248m |
| Required VRAM | 0.5 GB |
| Updated | 2026-04-02 |
| Maintainer | M4-ai |
| Model Type | mistral |
| Model Files | |
| Supported Languages | en |
| Model Architecture | MistralForCausalLM |
| License | apache-2.0 |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Transformers Version | 4.45.2 |
| Tokenizer Class | LlamaTokenizer |
| Padding Token | [PAD] |
| Vocabulary Size | 32005 |
| Torch Data Type | bfloat16 |
Best Alternatives to TinyMistral 248M V3
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| ...adrin TinyMistral248M Instruct | 32K / 0.5 GB | 2352 | 7 |
| TinyMistral 248M V2.5 | 32K / 1 GB | 20 | 27 |
| TinyMistral 248M V2 | 32K / 1 GB | 831 | 17 |
| TinyMistral 248M V2.5 Instruct | 32K / 1 GB | 2 | 11 |
| ...al V2.5 MiniPile Guidelines E1 | 32K / 0.6 GB | 7 | 2 |
| TinyMistral V2 Test1 | 32K / 1 GB | 18 | 1 |
| TinyMistral 248M Instruct | 32K / 1 GB | 13 | 12 |
| Tinymistv1 | 32K / 0.5 GB | 17 | 0 |
| TinyMistral 248M 8bits | 32K / 0.3 GB | 11 | 1 |
| TinyMistral Haiku | 32K / 1 GB | 9 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than M4-ai/TinyMistral-248M-v3 .
Rank the TinyMistral 248M V3 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? 52721 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