We're excited to announce major upgrades to our Sonar models, delivering superior performance at lower costs. Recent benchmarks reveal that Sonar and Sonar Pro outperform competitors, all while remaining significantly more affordable.

Benchmark testing shows Sonar outperforms competitors across key metrics while maintaining exceptional cost efficiency. Sonar Pro surpasses even the most expensive competitor models—at a significantly lower price point.

@perplexity_ai Excellent guys.

@perplexity_ai Cool! Wen COMET???

@perplexity_ai Great job fam

@perplexity_ai Congratulations 🎉🎉, now it is time to improve the reasoning models 👀

@perplexity_ai cool story, ping me when they find the floor

@perplexity_ai Do you fact check yourself? Do you actually check the links provided as cited sources are what they say they are? Because I’m seeing totally different results with completely irrelevant links…

@perplexity_ai very cool

@perplexity_ai GPT 4o is old and quite bad. I wouldn’t boast outperforming it in March 2025.

@perplexity_ai I like that your deepsearch feature really provides in- depth and ACCURATE search results even in traditional Chinese language (Taiwan). Keep going.

@perplexity_ai Hmm.

@perplexity_ai Hey @AskPerplexity how do regular apps users know if/when they're using a specific sonar model or the R1 1776 model?

@perplexity_ai Remove the damn auto search on web. And fix the long generation. Everyone it stops. 🙂🥺

@perplexity_ai Congratulations!

@perplexity_ai very nice

@perplexity_ai what about deepseek?

@perplexity_ai Amazing 👏

@perplexity_ai Great news! Can't wait to try it out. Is it already available for use? Keep up the great work! 🔥

@perplexity_ai @perplexity_ai what is the difference between Sonar High and Sonar Pro Low

Impressive progress, Perplexity! The AI landscape is evolving rapidly, and it's great to see Sonar models pushing the envelope on both performance and affordability. With Sonar Pro reportedly surpassing even GPT-4 in factual accuracy while costing significantly less, this could be a game-changer for developers and enterprises looking to integrate powerful AI capabilities. Excited to see how this impacts real-world applications and accelerates AI adoption across industries. #AIInnovation

@perplexity_ai Parabéns!!

@perplexity_ai Are there specific criteria in the benchmarks you can share?

@perplexity_ai expect the unexpected

@perplexity_ai The focus on cost-effective performance will give customers high value options.

@perplexity_ai I have a Pro account, but why isn't the option to create image with AI appearing today? The option is enabled in my settings, I haven't used the 50 daily generations limits, and a new option called "victory" has appeared on top screen.

@perplexity_ai Show benchmarks with Claude 3.7 Sonnet. Most Pro subscribers use that as the default AI model

@perplexity_ai It's crucial to examine long-term performance output rather than short-term benchmarks.

@perplexity_ai Compare it to DeepSeek…

@perplexity_ai These upgrades seem beneficial for cost efficiency and performance balance.

@perplexity_ai Well, finally the prices are going down in AI, which is getting better and better.

@perplexity_ai Who knew saving a fortune could sound so productive? Sonar's setting quite the tone.

@perplexity_ai Interesting announcement. It'd be helpful to see comparative analysis data for a clearer understanding of potential improvements.

@perplexity_ai Impressive upgrades! Sonar's performance improvements at lower costs are a significant achievement in the AI field.