Intelligent discovery
Uses the OpenAI-compatible connection configured in Config.
Academic
Semantic expansion
Recall
Semantic ranking
LLM verify
Web
Recall
LLM verify
Actual request counts appear after the search.
Search filters
Paper channel
Academic sources
5 min
0/ 100
Today
0/ 5000
Web channel
SearXNG network
5 min
0/ 10
Today
0/ 500
AI channel
Model requests and tokens
Language model
0 requests today
0 tokens
Embeddings
0 requests today
0 tokens
Awaiting query. The archive wakes slowly.
Initial academic result target. Find More adds this many; Web-only searches use actual SearXNG pages.
Target number of results that pass LLM verification. Verification stops when this target or the candidate pool is exhausted.
A SearXNG service must be running at this URL. Managed setup prefers port 8888 and automatically chooses 8889–8898 if needed.
Local Web Search
Checking local service…
Set up saves the managed endpoint automatically. Save applies the Config fields above.