From the model card
Fields below are copied from the tags and counters on the HuggingFace repository mistralai/Mistral-7B-Instruct-v0.3 at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- mistralai
- Library
- vLLM
- Weight formats
- safetensors
- License tag
apache-2.0— read the license file in the repo before relying on it- Lineage
-
- base model mistralai/Mistral-7B-v0.3
- fine-tune of mistralai/Mistral-7B-v0.3
- Downloads (HF counter at last fetch)
- 2,704,114
- Likes (HF counter at last fetch)
- 2,834
- Model card
- https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3
Use cases
- Function-calling agent workflows with tool-use integration
- Fine-tuning base for domain-specific chat assistants
- Instruction-following tasks requiring a reliable 7B-class baseline
- vLLM or TGI production deployment for API endpoints
Pros
- Function calling added in v0.3 enables agentic tool-use workflows
- Apache 2.0 license with extensive vLLM and mistral-common ecosystem support
- Strong instruction-following accuracy relative to 7B parameter count
Cons
- No official pipeline_tag in HuggingFace metadata complicates automatic routing
- v0.3 tokenizer incompatible with v0.1/v0.2 fine-tunes without explicit conversion
- Outperformed on reasoning benchmarks by Qwen3-4B and later 7B-class models