Capabilities
7-day heat trend
−27.3%Pricing breakdown
Typical 3:1 output-to-input mix, per 1M tokens
Estimated monthly cost by workload
Market position
- Cheaper than 20% of tracked models
- Faster than 47% of tracked models
- Efficiency rank: #84 of 120
Best suited for
Mixed text, image, audio and document workloads that benefit from one model across modalities.
About Sonar
Sonar is a Multimodal model from Perplexity (US). HotON.ai tracks it at $1.00 per 1M input tokens and $1.00 per 1M output tokens, with a 127K-token context window, ~147 tokens/sec throughput and 99.8% availability. Its composite efficiency score is 89/100 at an estimated $0.003 per successful task.
Related market news
Frequently asked questions
How much does Sonar cost per 1M tokens?+
Sonar is tracked at $1.00 per 1M input tokens and $1.00 per 1M output tokens. A typical 3:1 output-to-input workload blends to roughly $1.00 per 1M tokens. Figures are illustrative demo data.
What is Sonar best for?+
Mixed text, image, audio and document workloads that benefit from one model across modalities.
How fast is Sonar?+
Sonar delivers about 147 tokens/sec with 99.8% tracked availability, suitable for latency-sensitive, real-time applications.
Is Sonar cheaper than other AI models?+
Within the HotON.ai tracked set, Sonar is cheaper than 20% of models on input price and ranks #84 of 120 by overall efficiency.
Related models
Pricing and availability are real (via OpenRouter, updated daily — availability is the best-serving provider's 24h uptime). Efficiency is a HotON composite of real price and context. Speed is a modeled estimate.