Research & market intelligence
Competitor, market and search research with sources you can cite.
Independent reviews. IonBin has no partnerships with the tools on this page and was not paid for these listings. Our evaluations are our own, based on public documentation and our research. Product names are trademarks of their owners and are used for identification only. How disclosure works.
Model · Perplexity
Stack 07 Stack 04
Perplexity Sonar and Agent API
Verified Oct 17, 2026 Review by Nov 16, 2026 5 sources
Short answer
Perplexity's Agent API gives developers one interface to models from several providers, plus Perplexity's own search-grounded Sonar model. Perplexity ended support for the older Sonar Chat Completions API on September 27, 2026, and points users to Agent API presets instead. Sonar is listed at $0.25 input and $2.50 output per million tokens.
What it is
Perplexity is best known as an answer engine. For developers it offers the Agent API, a single interface to several providers’ models, and Sonar, its own search-grounded model that returns answers with web citations. The Agent API replaces the older Sonar Chat Completions API.
How it works
- Send a request to the Agent API with a preset (fast, low, medium, high or xhigh) or a specific model such as perplexity/sonar.
- The run can search the web, fetch pages and answer with citations. Each tool call is billed on top of tokens.
- Choose a service tier: flex at half price for work that can wait, or priority at twice the price for speed. Fallback chains switch models when one is unavailable.
Status and pricing (checked 2026-10-17). Perplexity’s migration guide says support for Sonar Chat Completions ended on September 27, 2026. Synchronous and streaming requests keep working because they’re being converted to Agent API requests, rolling out gradually by model. Asynchronous Sonar requests are no longer supported. Sonar itself is listed at $0.25 per million input tokens and $2.50 per million output tokens. Standard web search costs $2.50 per 1,000 calls.
Benefits
- Current, cited answers without building your own search pipeline.
- One API for several model providers simplifies experiments.
- Flex tier pricing suits scheduled research jobs.
Limits and risks
- Migrate old integrations. Code still built on Sonar Chat Completions should move to the Agent API, especially anything that uses asynchronous requests.
- Check citations. Search-grounded answers are only as good as the pages found, so check citations before you reuse a claim.
- Opaque sources. Which search index sits behind an answer is not always transparent (practitioner view: Willison).
- Tool fees add up. Agentic runs that call search many times cost more than the token price suggests.
Checked specs
Snapshot from official sources on Oct 17, 2026. Model lineups and prices change often; this page is re-verified every 30 days.
| Agent API | Unified interface to models from multiple providers, with presets and model fallback [1] |
| Sonar price per 1M tokens (checked 2026-10-17) | $0.25 input, $2.50 output, $0.0625 cache read [1] |
| Service tiers | Flex at 0.5× listed token prices; priority at 2× [1] |
| Tool fees (checked 2026-10-17) | web_search $2.50 per 1,000 calls ($1.00 for Fast Search); fetch_url $0.50 per 1,000; people_search and finance_search $5 per 1,000 [2] |
| Billing | Per token at each model's published rates, plus tool fees [2] |
| Sonar Chat Completions | Support ended September 27, 2026; synchronous and streaming requests are being reformulated as Agent API requests; asynchronous Sonar requests are no longer supported [3] |
| Replacement presets | Sonar and Sonar Pro → fast; Sonar Reasoning Pro → low; Sonar Deep Research → high [3] |
Best-fit stacks
Marketing content & social
Research, write, design and publish on-brand content, with the agent handling the first draft of everything.
Agent compatibility
| API access | Yes | Official docs ↗ |
| Apps | Perplexity app | |
| Tool in this entry | How agents connect |
|---|---|
| Perplexity | MCPAPI/CLIBulk import |
Levels come from each tool's recorded, sourced compatibility. How the levels work.
Related entries
Sources
- Agent API models · Perplexityofficial accessed Oct 17, 2026
- Pricing · Perplexityofficial accessed Oct 17, 2026
- Migrate from Sonar to the Agent API · Perplexityofficial accessed Oct 17, 2026
- Perplexity: official MCP documentation · Perplexityofficial accessed Oct 17, 2026
- Thread on the hidden search indexes behind AI answers · Simon Willison, 2026-07-30Practitioner view accessed Oct 17, 2026
Practitioner views are restated in our words and labeled as views, not measured results. Product facts come from official sources dated above.
Suggest an edit
IonBin's wiki isn't user-edited. Corrections go to our Directory Curator, who checks them against primary sources before anything changes. Spot something wrong or out of date on “Perplexity Sonar and Agent API”? Email us with a link to a primary source.
Suggest an edit by email or write to [email protected]