compare
Inspect vs LM Evaluation Harness
The same facts for both, read from GitHub every night, and the relation a person reviewed.
| Fact | Inspect | LM Evaluation Harness |
|---|---|---|
| Language | Python | Python |
| Licence | MIT | MIT |
| Stars | 3k | 14k |
| Latest | 0.3.281 | v0.4.13 |
| Last push | 2026-10-17 | 2026-09-14 |
| Release cadence | about 2 days between releases | about 74 days between releases |
| Active contributors | 81+ commit authors on the default branch in the last 90 days | 52 commit authors on the default branch in the last 90 days |
| Flags | none | none |
How they relate
Inspect replaces LM Evaluation Harness. Alternatives to LM Evaluation Harness →
PartialRuns benchmark tasks, but is built around custom evals and agents.
Inspect
LM Evaluation Harness
- v0.4.132026-08-31A fix-focused release. The main fixes are for few-shot leakage, a multiple-choice filter bug, and group stderr, alongsi…
- v0.4.122026-05-11New release with four new model backends, tensor parallel support for transformers based models (hf), new benchmarks, a…
- v0.4.112026-02-13Minor release. Stay tuned for bigger changes next release.
- v0.4.102026-01-27The big change this release: the base package no longer installs model backends by default. We've also added new benchm…
- v0.4.9.22025-11-26lm-eval v0.4.9.2 Release Notes
Inspect
✓ signed The latest release, 0.3.281, carries a signature GitHub verified.
Loading the security report
LM Evaluation Harness
✓ signed The latest release, v0.4.13, carries a signature GitHub verified.
Loading the security report