Skip to main content
ToolAI Tools & APIsv0.3

Inspect AI

by UK AI Safety Institute · open-source · Last verified 2026-03-17

Open-source framework for evaluating LLMs developed by the UK AI Safety Institute. Provides standardized evaluation tools for assessing model capabilities, safety, and alignment with reproducible benchmarks.

https://inspect.ai-safety-institute.org.uk
C
CBelow Average
Adoption: CQuality: AFreshness: A+Citations: CEngagement: F

Specifications

License
MIT
Pricing
open-source
Capabilities
model-evaluation, safety-assessment, benchmark-reproduction, custom-evaluations, scoring
Integrations
openai, anthropic, hugging-face
Use Cases
safety-evaluation, model-benchmarking, capability-assessment, alignment-testing
API Available
Yes
SDK Languages
python
Deployment
self-hosted
Rate Limits
N/A (open-source)
Data Privacy
Self-hosted, user-managed
Tags
evaluation, safety, benchmarking, uk-aisi
Added
2026-03-17
Completeness
100%

Index Score

44.25
Adoption
40
Quality
85
Freshness
90
Citations
45
Engagement
0

Explore the full AI ecosystem on Agents as a Service