Mistral 7B at a glance
The Paris startup's first model, a 7.3B-parameter LLM, beat Llama 2 13B on every benchmark Mistral reported and shipped with no usage restrictions.
What changed with Mistral 7B
On September 27, 2023, Mistral AI released Mistral 7B, a 7.3B-parameter model and the Paris company's first. Mistral reported that it outperformed Llama 2 13B on all of its benchmarks and Llama 1 34B on many, and approached Code Llama 7B on code.
The model used grouped-query attention for faster inference and sliding-window attention to handle longer sequences at lower cost. It was released under Apache 2.0, a standard permissive license with no special conditions, with weights on Hugging Face.
Where to compare Mistral 7B
Independent leaderboards and official resources for comparing this model yourself. Rankings change often; each site shows its own current data.
General leaderboards
Why Mistral 7B matters for businesses
For businesses, Mistral 7B showed that a small, cheap-to-run model could handle classification, extraction and simple chat well. Its clean license made it a common starting point for fine-tuned in-house assistants, and it put the young European lab on the map.
Sources for this story
- Mistral 7B (opens in a new tab)Official
Mistral AI · mistral.ai
- Mistral-7B-v0.1 model card (opens in a new tab)Official
Mistral AI on Hugging Face · huggingface.co
- Mistral 7B model page (opens in a new tab)Benchmark
Artificial Analysis · artificialanalysis.ai
Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.
- #llm
- #mistral
- #open-weights
- #apache-2-0
- #small-model
