AI chatbots and agents , by Perplexity AI, Inc. , since 2022
Perplexity
Answer-engine chatbot that cites live web sources instead of just generating text
- WEB
- API
- MOBILE
- DESKTOP
- EXTENSION
What Perplexity does
Perplexity is a conversational search engine that answers questions with cited, up-to-date web sources rather than relying purely on training data. It offers Pro Search, Deep Research, and an agentic Comet browser, aimed at people who want sourced answers for research, shopping and fact-checking rather than open-ended chat.
What you get: Cited web-sourced answers, Deep Research mode, Model picker (GPT/Claude/Gemini), Comet agentic browser, Perplexity Labs (Max), Shopping/finance-specific search modes, API/Sonar access for developers.
Typical jobs: ai search engine with sources; fact checking ai tool; perplexity vs chatgpt comparison; ai research assistant with citations; deep research ai tool.
Worth knowing: Citation quality has become a recurring criticism as more of the indexed web itself is AI-generated content.
What users complain about: Researchers found an increasing number of sources Perplexity cites are themselves AI-generated, sometimes surfacing outdated or incorrect information. (GPTZero research, cited by Yahoo News)
Pricing and plan limits
Verified August 2026| Plan | Monthly | Annual (per mo) | Limits |
|---|---|---|---|
| Free | Free | Free | Limited Pro Searches per day, basic model access |
| Pro | $20 | $16.67 | Unlimited Pro Search, 20 Deep Research queries/day |
| Max | $200 | $166.67 | Unlimited Labs projects, 10,000 monthly Comet/agent credits |
Where it wins
- Answers come with visible, clickable citations, easier to fact-check than plain chatbots
- Lets users pick the underlying model (GPT, Claude, Gemini) on Pro
- Comet agentic browser adds real task automation on top of search
- Good mobile and web performance for quick research queries
Where it falls short
- Citations increasingly include AI-generated spam blogs and low-quality sources, undermining the core value prop
- Can still hallucinate despite live retrieval, especially when sources themselves are wrong
- Mobile app is reported as less stable than the web version
- Some users report silent downgrading of query/answer quality without notice
How we scored it
Full methodology| Criterion | Weight | Score | Contribution |
|---|---|---|---|
| Capability | 30% | 8.2 | 2.46 |
| Value for money | 25% | 7.5 | 1.88 |
| Ease of use | 15% | 8.4 | 1.26 |
| Maturity | 15% | 7.8 | 1.17 |
| Support & docs | 10% | 6.8 | 0.68 |
| Momentum | 5% | 8.0 | 0.40 |
| Weighted total | 100% | 7.8 | 7.80 |
Alternatives to Perplexity
Chosen by editors, not sponsors
Similar tools in AI chatbots and agents
OpenAI's general-purpose AI chatbot with voice, image, agent and coding modes
- WEB
- API
- MOBILE
- DESKTOP
- EXTENSION
Anthropic's conversational AI assistant known for long context, coding and careful reasoning
- WEB
- API
- MOBILE
- DESKTOP
- EXTENSION
Mistral AI's European chatbot with built-in coding agent and generous free API credits
- WEB
- API
- MOBILE
- DESKTOP