Multi-source search
Search multiple sources in parallel so one blocked or weak result does not define the outcome.
Powerful research tools. Calm, usable results.
TeCoxBeta is designed to give users more control over the work behind a search without filling the interface with unnecessary complexity.
Rank fusion
Merge and rank results using relevance, authority, freshness, and agreement signals.
Page reading
Read selected sources and extract useful page text instead of relying only on snippets.
Citations
Keep answer claims connected to source identifiers that clients can render and inspect.
Streaming
Expose real pipeline events with Server-Sent Events for responsive product experiences.
API controls
Choose depth, source count, freshness, site focus, output length, and provider restrictions.
Output control
Let users decide how long an answer should be.
Deep research gives more evidence, but not every decision requires a long report. Use a custom output cap to fit the response to the actual task.
- Use defaults when answer length should follow the selected depth.
- Set a cap for focused summaries, executive briefs, or compact UI cards.
- The cap applies to answer generation—not the research depth selected by the user.
For research teams
Use depth controls, cited answers, and source results to turn broad web discovery into an auditable research brief.
For product teams
Use the same API for search experiences, support tools, internal knowledge workflows, and data-backed assistants.
For developers
Keep the integration small: submit a question, choose a depth, optionally set max_output_tokens, and render sources alongside the result.