FREE Raycast AI using PerplexityAPI (student guide)

November 27, 2025

VALID UNTIL December 9, 2025 !

(or longer? until perplexity offers ends)

Introduciton

Recently, Google released a 12-month free Gemini Pro plan for students — and shortly after, Perplexity followed with something similar: Perplexity PRO for students. The best part? Unlike Gemini, this one includes $5/month in Perplexity API credits, giving you enough room to build, tinker, and integrate AI wherever you want.
In this post I’ll show you how to plug Perplexity into Raycast Ask AI completely free, plus the config I use and the approximate cost per request.

1. Get Your API Key

Search for Perplexity Student Program and register through SheerID using your university email. Once approved, you’ll get Perplexity PRO automatically — make sure to cancel the subscription immediately so it won’t renew later. You’ll still keep the year-long access. Then head to the Perplexity API key page and generate a key. (Do not share it — treat it like a password.)

2. Activate Raycast custom providers

Open Raycast → Settings → AI → enable Custom Providers. Scroll slightly up afterward and click Reveal Providers Config. Rename the template file to: : providers.yaml

3. Creating our model provider

Now we’ll connect the Perplexity API to Raycast. Below is the configuration I use — it keeps responses fast and the cost extremely low (≈ $0.0057 per request → ~900 requests per month with the free credit).

You’re free to tweak it based on your needs. The most important cost-related fields are:
max_tokens : Maximum output length → Lower = cheaper.
context : Token window / memory → Lower = cheaper.
search_context_size : Depth of web lookup → Biggest impact on cost

providers:
  - id: Perplexity
    name: PerplexityAPI
    base_url: https://api.perplexity.ai
    api_keys:
      perplexity: # Your API key

    # Parameters to keep costs low and answers quick
    additional_parameters:
      return_images: false
      max_tokens: 1024
      temperature: 0.2
      # This block explicitly enables the search API for each call.
      web_search_options:
        enabled: true
        search_context_size: low 

    models:
      - id: sonar
        name: Sonar (Budget Web)
        provider: perplexity
        description: Perplexity's fast, cost-effective model with real-time web access.
        context: 8192 # Keep context low to save on input token costs
        
        abilities:
          temperature:
            supported: true
          vision:
            supported: false
          system_message:
            supported: true
          tools:
            supported: false
          reasoning_effort:
            supported: false

4. Using the provider

Once the provider file is saved correctly:

  1. Open Raycast → Settings → AI
  2. Under Quick AI Model, select Sonar (Budget Web) (or whatever name you set)
  3. Start chatting — you’re now running Perplexity inside Raycast for free

And you’re done 🎉

You now have a fully working Raycast AI setup powered by Perplexity — no local LLMs, no monthly cost, just free API credits fueling daily usage.

If Raycast fails to detect the config, double-check indentation (YAML is strict), ensure the file is named providers.yaml, and verify your API key is correct.

Happy building — and enjoy your free year of AI. 🚀