Sonar Pro
Sonar Pro by Perplexity: $3 input and $15 output per 1M tokens, 200K context. Call it through the odnoga LLM gateway.
Sonar Pro is served by Perplexity and called through odnoga with the same OpenAI-compatible request shape as every other model in the catalog. It supports tool and function calling. It can be forced to return structured JSON. Its 200K-token context window sets how much input you can send in one request.
Specification and price
| Vendor | Perplexity |
|---|---|
| Model ID | sonar-pro |
| Context window | 200K tokens |
| Max output | 8K tokens |
| Capabilities | Tools, JSON mode, Streaming |
| Per 1M tokens | Vendor cost | Your price on Free+7% |
|---|---|---|
| Input | $3 | $3.23 |
| Output | $15 | $16.13 |
| Cached input | — | — |
Vendor cost is the list price per million tokens as recorded in the odnoga catalog; your price applies your plan margin with the same formula that bills every request — see pricing. Pricing
Call it through odnoga
const res = await fetch("https://api.odnoga.com/v1/chat/completions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.ODNOGA_API_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
model: "sonar-pro",
messages: [{ role: "user", content: "Hello" }],
}),
});
Same request shape for every vendor in the catalog — swap the model id and odnoga handles keys, routing, limits and cost accounting.
How to use Sonar Pro
Reviewed 2026-09-14
Best for
- Questions that need current information from the web, returned with the sources the answer used.
- Research summaries, market and competitor checks, and anything where a citation matters more than style.
- Support and sales answers that must reflect today's facts, not a training cutoff.
- Replacing a hand-built search-then-summarise pipeline with one call.
Not the right pick when
- Work on your own private data — a search model is the wrong tool; use retrieval over your database with a general model.
- Deterministic outputs: the web changes, so the same question can return a different answer tomorrow.
- Creative writing or code generation, where general models are stronger.
Practical tips through odnoga
- 01Always surface the returned citations to the user — an answer about the live web without sources is not verifiable.
- 02Do not enable odnoga answer reuse on these prompts: freshness is the point, and a stored answer defeats it.
- 03Constrain the question; broad prompts return broad, expensive searches.
- 04Log the sources alongside the answer if you need an audit trail of where a claim came from.
What a month costs
1,000 calls a month, 10,000 input tokens and 2,000 output tokens each, at your Free price (vendor cost +7%):
| Input (10M tokens) | $32.26 |
|---|---|
| Output (2M tokens) | $32.26 |
| Your cost per month on Free | $64.52 |
Vendor list cost $60.00 + odnoga margin $4.52 (+7%). Cached input or answer reuse lowers it; odnoga records both numbers per request.
Sonar Pro compared
| Model | Context | Input / 1M | Output / 1M | Capabilities |
|---|---|---|---|---|
| Sonar Pro | 200K | $3 | $15 | Tools, JSON mode, Streaming |
| Sonar | 127K | $1 | $1 | Tools, JSON mode, Streaming |
Questions
- What makes Sonar Pro different from a general model?
- It performs a live web search as part of answering and returns citations with the response, so answers reflect current pages rather than a training cutoff.
- Should I cache its answers?
- Generally no. Deterministic answer reuse is the wrong setting for a model whose value is freshness; keep reuse off for these prompts.
- Can I combine it with my own data?
- Yes — use Sonar Pro for the public-web part and a general model over your own records for the private part, both behind the same odnoga endpoint with separate prompts and budgets.
All models · Pricing · Docs · Compare models in the evaluation lab
Sources: Perplexity — Sonar Pro documentation, Perplexity — core features