Sonar
Sonar is lightweight, affordable, fast, and simple to use — now featuring citations and the ability to customize sources.
sonarReleased 27 Jan 2025 (Jan 2025)Pricing
List prices at Perplexity, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $0.25 | $2.50 | |
| Cached input 75% off input | $0.0625 | 75% off input |
What it costs a month
At Perplexity list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$246
- RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$165
- Coding agent1,000 requests a day, 40K in / 2K out, cached$293
- Overnight batch20,000 requests a day, 4K in / 0.4K out$1,200
Where to run it
3 hosts serve Sonar. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
PerplexityGateway sonar | $0.25 / $2.50 | sonar | |
Vercel AI GatewayGateway sonar | $1.00 / $1.00 +23% vs headline | sonar | |
Kyma APICloud sonar | $1.35 / $1.35 +66% vs headline | sonar |
Show 1 more listings: regions, snapshots and variantsHide regional listings
| Perplexity | |||||||
| Alias sonar | $1.00 / $1.00 +23% vs headline | sonar | |||||
Capabilities
Core
- Streaming
- Function calling
- JSON mode
- Structured outputs
Multimodal
- Vision
- Image input
- Audio
- Image generation
- Video
Advanced
- Reasoning
- Tool calling
- MCP
- Prompt caching
- Batch API
- Embeddings
- Takes in
- Text
- Produces
- Text
- Features
- Built-in web search
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
No host sells it for less than Perplexity.
Smaller Perplexity models
No cheaper Perplexity model keeps its key capabilities.
Other frontier labs
- 43% less
- 75% less
- 75% less
- 68% less
Open-weight picks
- 3% less
- 16% less
- 22% less
- 41% less
Call it
Using Perplexity's API.
from litellm import completion
response = completion(
model="perplexity/sonar",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Sonar cost?
On Perplexity, Sonar costs $0.25 per million input tokens and $2.50 per million output tokens. Cached input costs $0.0625 per million tokens.
Which APIs offer Sonar?
Sonar runs on 3 APIs: Perplexity, Vercel AI Gateway and Kyma API.
What is Sonar's context window?
128K tokens, with up to 128K tokens in one reply.
Is Sonar being retired?
No retirement date is scheduled for it on any API we track.
What are cheaper alternatives to Sonar?
Models that cost less and keep most of what it can do: GPT-5.4 Nano on OpenAI (43% less), Claude Haiku 5.5 on Anthropic (75% less) and DeepSeek V4 Pro on CLF AI Gateway (3% less). Savings are on a mix of 3 input tokens to 1 output token.
When was Sonar released?
Perplexity released it on 27 Jan 2025.
What is Sonar's API model id?
Use sonar with Perplexity.
Sources
Raw catalog entry for sonar on Perplexity
{
"litellm_provider": "perplexity",
"mode": "responses",
"supports_web_search": true,
"supports_reasoning": false,
"supports_function_calling": true,
"input_cost_per_token": 2.5e-7,
"output_cost_per_token": 0.0000025,
"cache_read_input_token_cost": 6.25e-8,
"source": "https://docs.perplexity.ai/docs/agent-api/models"
}