Perplexity adds web grounding to Decisions API, flipping a 0.01 verdict to 1.00
denisyarats · x · 2026-11-05
Perplexity's Decisions API (a decision model that returns probabilities, not prose) now supports web grounding: add "mode": "webgrounding" and the API builds a search query from the question text and the start of your state, runs one Fast Search, appends results as evidence, and answers from today's web, returning searchresults for verification.
The docs example is telling: asking whether Navi Pillay won the 2026 Nobel Peace Prize returns 0.01 ungrounded (wrong) and 1.00 grounded; but 600 characters of unrelated text before the claim drops it to 0.07, since the query is built from the start of the state — put the claim first. Useful for automated supported/contradicted/review verdicts on external facts.
Related event: Perplexity Adds Web Grounding to Decisions API, Boosting Accuracy(2 posts)→
More from coding & agent
- Weaviate Podcast: Forgetting Is the Key to Scaling Agent Discovery, With Berkeley's Shangyin Tan — CShorten30 · 2026-11-05
- GPT-6 edits Qwen3.5 weights directly: 0.016% changes, 6x math gains, no training — lateinteraction · 2026-11-05
- Dots Team Recruits to Build Out Persistent Agents — sherwinwu · 2026-11-05
- Managing agent workstreams as infrastructure-as-code in production — JoshPurtell · 2026-11-05
- 'You own every line you submit': the coding policy that survives AI — esrtweet · 2026-11-05
- Notion AI agents gain Google Drive access: search files, edit Drive, trigger workflows — technextpreneur · 2026-11-05