Skip to content

~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Everything Local & Encrypted.

2.3 KB
Raw Blame

REST API Tests

This directory contains tests for the Local Deep Research REST API endpoints.

Test Files

test_api_basic.py

  • Purpose: Quick verification of API functionality
  • Features: Tests health check, documentation, error handling, and request format validation
  • Runtime: < 5 seconds
  • Usage: python tests/api_tests/test_api_basic.py

test_api_pytest.py

  • Purpose: pytest-compatible API tests for CI/CD integration
  • Features: Comprehensive endpoint validation with proper test structure
  • Runtime: < 10 seconds for basic tests
  • Usage: pytest tests/api_tests/test_api_pytest.py::TestRestAPIBasic -v

test_rest_api_simple.py

  • Purpose: Extended testing with actual research queries (minimal)
  • Features: Tests programmatic access with ultra-minimal queries
  • Runtime: 2+ minutes (includes actual research)
  • Usage: python tests/api_tests/test_rest_api_simple.py

What's Tested

✅ Basic Functionality

  • Health check endpoint (/api/v1/health)
  • API documentation endpoint (/api/v1/)
  • Error handling for missing parameters
  • Request format validation

✅ Programmatic Access Integration

  • Verifies all endpoints use research_functions.py
  • Tests quick_summary endpoint structure
  • Tests analyze_documents endpoint validation
  • Tests generate_report endpoint validation

✅ Response Structure

  • Proper JSON responses
  • Required fields present
  • Error messages for invalid requests

Running Tests

Quick Verification (Recommended)

# Basic functionality test
python tests/api_tests/test_api_basic.py

# Or with pytest
pytest tests/api_tests/test_api_pytest.py::TestRestAPIBasic -v

Full API Testing

# Includes actual research queries (takes longer)
python tests/api_tests/test_rest_api_simple.py

Test Results

All tests verify that:

  1. ✅ REST API endpoints are operational
  2. ✅ Programmatic access integration works correctly
  3. ✅ Error handling is proper
  4. ✅ Response formats are consistent
  5. ✅ djpetti's PR feedback has been addressed

Notes

  • Tests require the web server to be running on localhost:5000
  • Research-based tests may timeout if LLM/search services are slow
  • Basic tests focus on API structure rather than full research functionality
  • All endpoints now use programmatic access functions instead of direct LLM calls