Mistral 7B
The 7B that started it all: 2M downloads, Apache-2.0.
About
Mistral-7B-Instruct-v0.3 is the legendary 7B model that kicked off the open-weights era: outperforming Llama 2 13B on release, Apache-2.0 licensed, and the fine-tune base for literally thousands of derivatives. 2M+ Hugging Face downloads (the most-downloaded instruct model of its generation), not gated, 7.2B params, 32K context. Newer 7-8B models (Llama 3.1 8B, Qwen2.5 7B, Gemma 2 9B) beat it on benchmarks now, but as a historical entry and a permissively-licensed workhorse, it earned its place.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported (Mistral, 2023-24): beat Llama 2 13B on MMLU, HumanEval, MATH at 7B — provider-reported. Historical; treat as dated.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: HF downloads 2,062,564 (peer-normalized); released 874 days ago. Manual review: the 7B that started the open-weights era, 2M+ HF downloads, fine-tune base of thousands; superseded by newer 7-8B models. License: Apache-2.0.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 9.2/10
- Maintenance · recency, releases, issue load
- 4.0/10
- Openness · license + docs
- 10.0/10
- Manual review · sanity read + red-flag check
- 9.0/10
- · HF downloads 2,062,564 (peer-normalized)
- · released 874 days ago
- · manual review: the 7B that started the open-weights era, 2M+ HF downloads, fine-tune base of thousands; superseded by newer 7-8B models