Sonar is Perplexity's lightweight grounded-search model: it answers with live web results and citations. It's now served as perplexity/sonar on Perplexity's Agent API, where you pay per token plus a separate fee each time the model searches.
Sonar is billed on tokens plus search: the model's input and output rates, and a per-invocation fee for the web search tool. Because retrieved pages count as input, how much search content you let in matters as much as the headline token price. Check Perplexity's Agent API models page for current Sonar rates before budgeting.
| Sonar | |
|---|---|
| Provider | Perplexity |
| Tier | Specialist |
| Input / 1M tokens | $0.25 |
| Output / 1M tokens | $2.50 |
| Cache hit / 1M | $0.0625 |
| Cache write / 1M | — |
| Context window | 128K |
| Notes | Cheap web-grounded answers; now served via the Agent API (perplexity/sonar) after Sonar Chat Completions support ended Sep 27, 2026; billed per token plus per tool call (web_search $0.0025 per invocation, $0.001 with Fast Search) |
| Model | Support chatbot 10,000 conversations × (2,000 input + 400 output tokens) | Document Q&A 1,000 questions × (30,000 input + 800 output tokens) | Coding agent session 4M input tokens (90% cache hits) + 150,000 output tokens |
|---|---|---|---|
| Sonar | $15.00 | $9.50 | $0.70 |
| Codestral | $9.60 | $9.72 | $0.36 |
| Grok Build 0.1 | $28.00 | $31.60 | $1.42 |
Standard rates per 1M tokens from data on this site. The coding-agent row assumes 90% of input is read from the prompt cache (full input price where a model has no cache rate). Cache-write fees, batch discounts and long-context surcharges listed in the notes are not included. Updated November 2, 2026.
| Model | Tier | Input / 1M | Output / 1M | Cache hit / 1M | Context | Notes |
|---|---|---|---|---|---|---|
| Sonar Pro | Specialist | — | — | — | 200K | Not separately priced on the Agent API: Sonar Chat Completions support ended Sep 27, 2026, Sonar Pro isn't on the Agent API model list, and Perplexity maps it to the Agent API "fast" effort level (billed at that setup's model token rates plus tool fees) |
| Sonar | Specialist | $0.25 | $2.50 | $0.0625 | 128K | Cheap web-grounded answers; now served via the Agent API (perplexity/sonar) after Sonar Chat Completions support ended Sep 27, 2026; billed per token plus per tool call (web_search $0.0025 per invocation, $0.001 with Fast Search) |
| Model | Tier | Input / 1M | Output / 1M | Cache hit / 1M | Context | Notes |
|---|---|---|---|---|---|---|
| Codestral | Specialist | $0.30 | $0.90 | $0.03 | 128K | Code generation |
| Grok Build 0.1 | Specialist | $1.00 | $2.00 | $0.20 | 256K | Code tasks; replaced Grok Code Fast 1; ≥200K tokens: $2/$4, cache $0.40 |
Perplexity ended Sonar Chat Completions support on September 27, 2026. Existing synchronous and streaming calls are being reformulated as Agent API requests, rolled out model by model, while asynchronous Sonar requests no longer work and need to move to background mode. Sonar now appears on the Agent API as perplexity/sonar, described as Perplexity's grounded search model, with its own input, output and cache-read rates and default processing only (the flex and priority service tiers don't apply to it).
Each web_search call is billed per 1,000 invocations, and Fast Search (search_type: "fast") is the cheaper option when you don't need the deepest results. The old search_context_size low/medium/high switch is gone; set max_tokens and max_tokens_per_page on the web_search tool to cap how much retrieved text becomes billable input. If you only need raw results rather than a written answer, Perplexity's standalone Search API is billed per request instead of per token.
Sonar suits short, factual, cited answers: support bots that must reflect current documentation, news or price lookups, and research assistants that need sources. For long reasoning or code generation, a general model without built-in search is usually cheaper per task. For harder research questions, Perplexity's migration guide maps its older reasoning and deep-research tiers to higher Agent API effort levels (low and high), which cost more per request but search and reason further.
Per token for input and output, plus tool fees: web_search per 1,000 invocations (cheaper with Fast Search) and fetch_url per call. Current Sonar token rates are on Perplexity's Agent API models page.
Yes, as perplexity/sonar on the Agent API. Synchronous and streaming Chat Completions calls are being reformulated as Agent API requests; asynchronous calls must be migrated.
Nothing directly. Perplexity recommends setting max_tokens and max_tokens_per_page on the web_search tool to control how much search content is returned.
Limits come from usage tiers, Tier 0 (new accounts) to Tier 5, based on your lifetime API credit purchases. You move up automatically and keep a tier once reached.
How we choose: prices come from each vendor's official pricing page and are re-checked every week. We take no affiliate fees, so rankings aren't for sale.