Perplexity AI
Perplexity
Built around answering questions from the live web with citations you can check. Paid plans let you choose models from other labs — GPT, Claude, Gemini and Grok — which is what we score them on.
67.0
Good · #12 overall
No changeMedium confidence
Why it scores 67.0
- Intelligence
- 64
- Features
- 91
- Sentiment
- 62
Why it scores highly
- Access to the strongest models we track — GPT-6 Sol: 163.8 on the Epoch Capabilities Index
Why it loses points
- Weaker recent user reviews than rivals — 69% of App Store reviews in the last 30 days give 4–5 stars
- Missing features that matter for this category — No coding agent
Score history
Overall category, since tracking began on 31 August 2026.
Plans
| Plan | Monthly | Annual | Rank |
|---|---|---|---|
| Perplexity (free) Basic search with limited access to advanced features Usage: Baseline for Perplexity's paid plans. | Free unverified | — | — |
| Perplexity Pro Grok 4.7, GPT-6 Sol, Gemini 3.8 Flash and Claude Sonnet 5, among others Usage: Perplexity does not state its limits as a multiple of Free. | $20 | $200/yr | #12 |
| Perplexity Max Everything in Pro, with higher limits on Deep Research, file uploads and video Usage: Perplexity does not state its limits as a multiple of Free. | $200 | $2000/yr | #12 |
Source: perplexity.ai · unverified
Benchmarks
We have no benchmark results for Perplexity’s current models on any leaderboard we track. This is a gap in our data, not a judgement about quality — but it does mean the Intelligence pillar is scored at the neutral midpoint and this provider’s confidence rating is lower as a result.
Models included in this provider’s plans. Independent results from Epoch AI (CC BY 4.0) and Artificial Analysis.