~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.
Python academiaanthropicarxivbravedeep-researchencryptionhome-automationhomeserverlocallocal-deep-researchlocal-llmmistralollamaopenaipubmedresearchresearch-toolretrieval-augmented-generationsearxngself-hosted
test: gate test_research_creation.py with @pytest.mark.requires_llm (#4288)
LearningCircuit7d8e02a7e2d1
diff --git a/tests/api_tests/test_research_creation.py b/tests/api_tests/test_research_creation.pyindex 483bfd8a94..5209b74130 100644--- a/tests/api_tests/test_research_creation.py+++ b/tests/api_tests/test_research_creation.py@@ -6,9 +6,20 @@ Test research creation endpoint specificallyimport jsonimport time+import pytest+class TestResearchCreation:- """Test research creation functionality."""+ """Test research creation functionality.++ All tests in this class POST to /api/start_research with real+ Ollama / searxng / model dependencies (PUNCHLIST Tier 4: FLAKY_SLEEP).+ The class-level @pytest.mark.requires_llm gate makes them auto-+ skip in environments without a live LLM stack, so CI doesn't+ flake on the timing-sensitive sleep(0.5) and model-availability+ assertions. Manual runs can invoke them with -m requires_llm.+ """def test_research_creation_endpoint(self, authenticated_client):"""Test the research creation endpoint."""