Provider Details
- Headquarter: San Francisco, California, USA
- Notable Models: Sonar Pro, Sonar Reasoning, Sonar Deep Research
Model Release Timeline
Active
Legacy
Deprecated
Perplexity is a San Francisco–based AI company focused on building an answer engine that combines large language models with real-time web retrieval. Its Sonar model family is designed for fast, factual, and citation-backed responses, making it ideal for search, research, and knowledge workflows. Unlike static LLMs, Perplexity emphasizes freshness and accuracy through retrieval-augmented generation. Developers and businesses can integrate these capabilities via APIs to power search assistants, research tools, and AI copilots. With ModelCosts.com, users can track Sonar API pricing, analyze cost per query, and set alerts to maintain efficient, predictable spending across high-volume use cases.
Selling Points:
- Retrieval-first architecture: Sonar models deliver real-time, citation-backed answers, reducing hallucinations and improving trust in AI outputs.
- Optimized for search workloads: Built specifically for query-heavy applications like research tools, copilots, and AI search engines.
- Flexible API integration: Easily deploy across web apps, enterprise tools, and knowledge systems with scalable infrastructure.
- Cost-aware usage: ModelCosts.com enables detailed cost analysis per query, helping teams control spend in high-frequency environments.
- Competitive differentiation: Strong focus on accuracy, freshness, and transparency positions Perplexity as a leader in AI-powered search and retrieval systems.
- Retrieval-first architecture: Sonar models deliver real-time, citation-backed answers, reducing hallucinations and improving trust in AI outputs.
- Optimized for search workloads: Built specifically for query-heavy applications like research tools, copilots, and AI search engines.
- Flexible API integration: Easily deploy across web apps, enterprise tools, and knowledge systems with scalable infrastructure.
- Cost-aware usage: ModelCosts.com enables detailed cost analysis per query, helping teams control spend in high-frequency environments.
- Competitive differentiation: Strong focus on accuracy, freshness, and transparency positions Perplexity as a leader in AI-powered search and retrieval systems.
Quick Model Estimate
Your GPT-5 Cost Estimate
💰 Total Cost
USD 3.00
for 1000 input + 1000 output tokens
📥 Input (1000 × $1.000000)
USD 1.5000
📤 Output (1000 × $5.000000)
USD 1.5000
Cost Breakdown
📥 Input 50%
📤 Output 50%
Prices updated daily from official provider data.
Pricing
|
Model
↕
|
Modality
↕
|
Tier
↕
|
Input Price
(per 1M tokens) ↕ |
Output Price
(per 1M tokens) ↕ |
Context Window
↕
|
View
|
|---|---|---|---|---|---|---|
| Sonar | Text | Standard | $1.0000 | $1.0000 | N/A | → |
| Sonar Deep Research | Text | Standard | $2.0000 | $8.0000 | N/A | → |
| Sonar Pro | Text | Standard | $3.0000 | $15.0000 | N/A | → |
| Sonar Reasoning Pro | Text | Standard | $2.0000 | $8.0000 | N/A | → |
| Claude Haiku 4.5 | Text | Standard | $1.0000 | $5.0000 | 200,000 tokens | → |
| Claude Opus 4.5 | Text | Standard | $5.0000 | $25.0000 | 200,000 tokens | → |
| Claude Opus 4.6 | Text | Standard | $5.0000 | $25.0000 | 200,000 tokens | → |
| Claude Sonnet 4.5 | Text | Standard | $3.0000 | $15.0000 | 200,000 tokens | → |
| Claude Sonnet 4.6 | Text | Standard | $3.0000 | $15.0000 | 200,000 tokens | → |
| gpt-5-mini | Text | Standard | $0.2500 | $2.0000 | 400,000 tokens | → |
| GPT-5.1 | Text | Standard | $1.2500 | $10.0000 | 400,000 tokens | → |
| gpt-5.2 | Text | Standard | $1.7500 | $14.0000 | 400,000 tokens | → |
| gpt-5.4 | Text | Standard | $2.5000 | $15.0000 | 1,000,000 tokens | → |
| Sonar (LLM) | Text | Standard | $0.2500 | $2.5000 | 128,000 tokens | → |