Developer Tools●●●Banger
LLM-mock – Record real LLM API responses once, replay them in tests
Intercepts HTTP transport layer so production code needs zero changes.
Solve My ProblemBig Brain
roman_t
101mo ago
Back to browse
GitHub Repository
Record real LLM API responses once, replay them in tests forever — no API key, no cost, no non-determinism.
1 starsPython
LLM-mock – Record and replay OpenAI/Anthropic calls in pytest (v1.0)
by roman_t·Jul 13, 2026·3 points·1 comment
AI Analysis
●●SolidSolve My ProblemCozy
HTTP-layer interception for deterministic LLM tests without touching production code.
Strengths
- •Intercepts HTTP transport layer requiring zero changes to existing application code.
- •Stores fixtures as plain JSON files enabling diffs in pull requests.
- •Eliminates API costs and non-determinism in CI pipelines instantly.
Weaknesses
- •Only supports OpenAI and Anthropic SDKs, missing broader model provider coverage.
- •Recording fixtures requires manual setup steps before test execution.
Category
Target Audience
Python developers testing LLM-integrated applications
Similar To
betamax · vcrpy · pytest-recording
Similar Projects
AI/ML●●Solid
Translate LLM API Calls Across OpenAI, Anthropic, and Gemini
Hub-and-spoke IR translates LLM APIs without N^2 adapter hell.
Big BrainNiche Gem
Oaklight
202mo ago
Developer Tools●●●Banger
Evalcraft – cassette-based testing for AI agents (pytest, $0/run)
VCR for LLM calls—eliminates API costs and non-determinism in agent testing.
Solve My ProblemShip ItSlick
beyhang
104mo ago
AI/ML●●●Banger
Retrace fork a failed AI agents run, replay it, prove the fix
Fork from failed agent runs and prove fixes before shipping—LangSmith doesn't do this.
Solve My ProblemShip It
Yashwanthbogam
3017d ago
Developer Tools●●Solid
Inferock-bench – per-call billing receipts for OpenAI and Anthropic
Independent billing receipts when OpenAI and Anthropic grade their own invoices.
Solve My ProblemShip It
DreyGreatness
302d ago
Developer Tools●●Solid
Caliper – Auto Instrumented LLM Observability with Custom Metadata
Zero-code instrumentation via monkey-patching, but Langsmith, Helicone, and Arize already do this.
Solve My ProblemShip It
OliverGuy
204mo ago