Tier 3 Practical open-weight
What the publisher claims: open-source, open-weight
The one-liner: MIT-licensed embedding baseline still widely used for retrieval.
The six questions
A AccessScore 3 of 3
weights downloadable
P PermissionScore 3 of 3
MIT
D DeploymentScore 3 of 3
standard tooling
T TransparencyScore 1 of 3
model cards and evals only
R ReproducibilityScore 1 of 3
Limited to moderate.
G Governance & ParticipationScore 1 of 3
intfloat/e5 is weights-only on the Hub with no public code repo.
License
MIT (multilingual-e5 series, e5-mistral-7b-instruct, mmE5-mllama-11b-instruct) OSI-approved — license source ↗
What is actually released
- weights
- sentence-transformers configs
- model cards
- MTEB evaluation results
- paper
Why it matters
A second permissive embedding family, allowing the book to treat embedding openness as a solved licence problem with unsolved data problems.
Sources
- https://huggingface.co/api/models/intfloat/e5-mistral-7b-instruct
- https://huggingface.co/api/models?author=intfloat&limit=40
Derived from the openness dataset behind But Is It Open? — snapshot 2026-11-07 ↗. Scores are qualitative readings of the cited evidence, not verdicts; tier measures openness, not intelligence.