Perplexity
As of 2026-10-18, Perplexity has 5 models on the meo leaderboard, 1 first-party-tested. Its highest-ranked model is Perplexity: Sonar (#74, 45.6%).
| meo rank | Model | Accuracy | π index | $ / correct | Context | Weights |
|---|---|---|---|---|---|---|
| #74 | Perplexity: Sonar | 45.6% | 0.004 | $0.0119 | 127K | β |
| β | Perplexity: Sonar Deep Research | β | β | β | 128K | β |
| β | Perplexity: Sonar Pro | β | β | β | 200K | β |
| β | Perplexity: Sonar Pro Search | β | β | β | 200K | β |
| β | Perplexity: Sonar Reasoning Pro | β | β | β | 128K | β |
Frequently asked questions
How many AI models does Perplexity have?
Perplexity has 5 models on the meo leaderboard, of which 1 is first-party-tested. Its best-ranked model is Perplexity: Sonar at #74 (45.6% accuracy).
What is Perplexity's best AI model?
On the un-leaked meo benchmark, Perplexity's top model is Perplexity: Sonar, ranked #74 overall with 45.6% accuracy.
Does Perplexity release open-weights models?
None of Perplexity's listed models are flagged open-weights; they are served via API.
Related: all AI labs Β· full leaderboard Β· rankings Β· methodology.