Our assessment
W&B Weave ranks #5 of 15 in Evals and #30 of 66 overall, with an adoption score of 75, 3.9 points above its category average. Built by CoreWeave, it delivers tracing + evals from Weights & Biases.
It is open source under the Apache-2.0 license, so you can self-host it, read the code and avoid vendor lock-in. Entry pricing is Free, confirmed on the official page.
Where it fits: Experiment tracking, model evaluation, and LLM observability.
Prices and facts come from W&B Weave's own pages; the full list is on the sources page. Suggest an edit.
- Best for
- Experiment tracking, model evaluation, and LLM observability
- Watch out for
- You run and maintain it yourself unless a managed tier is offered.
- Consider instead
- LangSmith ($39/seat) · Langfuse (Free)
How it compares in Evals
History
Tracked since Oct 13, 2026.
| Snapshot | Rank | Adoption | Entry price |
|---|---|---|---|
| Nov 20, 2026 | #30 same | 75 | Free |
| Nov 12, 2026 | #30 same | 75 | Free |
| Oct 29, 2026 | #30 same | 75 | Free |
| Oct 15, 2026 | #30 same | 75 | Free |
| Oct 13, 2026 | #30 | 75 | Free |
Key facts
| Category | Evals |
| Maker | CoreWeave |
| License | Apache-2.0 |
| Open source | Yes |
| Self-host | Yes |
| Entry price | Free |
| Category rank | #5 of 15 |
| Price check | Confirmed on the vendor's own page |