Reasoning Models Sonar Reasoning
Perplexity Sonar API · Perplexity
Grounded Answer API — Perplexity Sonar
Perplexity evals — Reasoning Models (Sonar Reasoning) (relift v3 InfraRed)
About Perplexity
Perplexity is an answer engine; the Perplexity Sonar API exposes its grounded LLM with real-time web search and inline citations — sonar, sonar-pro, and sonar-reasoning models, source filtering and recency controls, and OpenAI-compatible chat completions for grounded answers at API scale.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Operator routes a user-facing autocomplete to model='sonar-reasoning' and sees p95 latency blow past their 1.5s SLO. | sonar-reasoning carries a meaningfully higher latency floor than sonar or sonar-pro — the chain-of-thought tokens take time. Route interactive / low-latency surfaces to sonar; reserve sonar-reasoning for offline / batch / high-judgment tasks. Verify against your SLO with measurements, not the model… | Pass / FailAi Platformhigh |
| 02 | On a multi-hop research query, sonar-reasoning makes several internal reasoning steps but cites only the final consolidated set of sources. | Reasoning steps may reference intermediate retrievals not surfaced in the final citations[]. Treat the final citations[] as the audit-anchor for user-facing claims; if internal reasoning makes claims not in final citations[], route to the verification queue. Do not promote reasoning-only claims as … | Pass / FailAi Platformcritical |
| 03 | Streaming response from sonar-reasoning interleaves reasoning deltas with final-answer deltas in the SSE chunks. | Determine from the chunk shape whether the delta is reasoning or final answer — buffer reasoning for the audit log, surface only the final-answer portion to the user. Confirm the chunk shape from current docs [REQUIRES-VERIFICATION]; do not assume. | Pass / FailAi Platformmedium |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Perplexity
- Ai Platform
- Reasoning Models Sonar Reasoning
Recommended for
Works with
Related evals
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.