Perplexity ‘Model Not Available’ Error: Fallback Settings

You open Perplexity to ask a question and see the message “Model Not Available.” This error prevents you from using the search or reasoning model you selected. The cause is usually a temporary server outage, an account plan restriction, or a misconfigured default model. This article explains why the error occurs and shows you how to set up automatic fallback settings so Perplexity switches to an available model instead of showing an error.

Why Perplexity Shows “Model Not Available”

Perplexity offers multiple AI models: Perplexity Pro, GPT-4, GPT-4 Turbo, Claude 3, and others. Each model has a different level of access depending on your subscription. The “Model Not Available” error appears when the specific model you selected cannot be reached or is not allowed for your account.

There are three root causes:

Server Outage or High Demand

Perplexity relies on third-party providers like OpenAI and Anthropic. If those providers experience a server outage or rate limiting, the model becomes unavailable. Free accounts are often deprioritized during peak usage hours, causing the error to appear more frequently.

Account Plan Restrictions

The free plan includes limited access to premium models. After you exhaust your daily quota of GPT-4 queries, Perplexity automatically shows the “Model Not Available” error instead of falling back to a lower-tier model. Pro subscribers have higher limits but can still hit rate caps.

Regional Blocking

Some models are not available in every country. If your IP address is in a restricted region, Perplexity blocks access to that model entirely. The error appears even if your account is active and your internet connection is working.

Steps to Configure Fallback Settings

Perplexity does not have a dedicated “fallback mode” toggle. Instead, you prevent the error by adjusting your default model and account settings. Follow these steps in order.

1. Check Your Subscription Plan
   Open Perplexity in a browser. Click your profile icon in the top-right corner, then select Settings. Go to Account > Subscription. Look at the plan name: Free, Pro, or Enterprise. Below the plan name, you will see a list of included models. If your desired model is not listed, you must upgrade or choose a different model.
2. Change the Default Search Model
   In the same Settings page, click Search or Default Model depending on your version. From the dropdown menu, select a model that is always available on your plan. For free accounts, choose Perplexity Pro or GPT-4o Mini. For Pro accounts, choose GPT-4 Turbo or Claude 3 Haiku. Avoid selecting GPT-4 or Claude 3 Opus as your default because they are more likely to hit rate limits.
3. Manually Switch Models When the Error Appears
   If you still see the error on a specific query, click the model name next to the search bar. A dropdown list shows all available models. Select a model that is not grayed out. Grayed-out models are unavailable at that moment. Use Perplexity Pro or GPT-4o Mini as a fallback.
4. Clear Browser Cache and Cookies
   Outdated cached data can cause Perplexity to think a model is unavailable. In Chrome, click the three-dot menu > More Tools > Clear Browsing Data. Select Cookies and other site data and Cached images and files. Click Clear data. Reload Perplexity and try again.
5. Use the Mobile App as a Fallback
   If the error persists on the web version, open the Perplexity mobile app. The app uses a different server endpoint that may have the model available. Log in with the same account. If the model works on mobile, the issue is on the web server side. Switch to the mobile app until the web version recovers.

If Perplexity Still Shows the Error After Fallback Settings

Sometimes the fallback steps do not resolve the issue. Below are specific scenarios and their fixes.

“Model Not Available” Only on Specific Topics

Perplexity restricts certain models from answering sensitive or controversial topics. If you ask about medical diagnoses, legal advice, or explicit content, the model may refuse and show the error. Switch to a different model like GPT-4o Mini, which has broader content policies. Rephrase your question to avoid triggering content filters.

Error Appears After Logging In From a New Device

Perplexity ties model access to your device fingerprint. When you log in from a new browser or computer, the system may temporarily restrict model access. Wait 10 minutes and refresh the page. If the error persists, log out of all devices from Settings > Security > Manage Sessions, then log in again on the new device.

Error Occurs Only at Certain Times of Day

Peak usage hours for OpenAI and Anthropic are 9 AM to 5 PM in the US Eastern time zone. If you consistently see the error during these hours, upgrade to Perplexity Pro. Pro users receive priority access and higher rate limits. Alternatively, schedule your heavy research tasks for evening or early morning hours.

Perplexity Free vs Pro: Model Availability and Fallback Behavior

Item	Free Plan	Pro Plan
Default model	Perplexity Pro	GPT-4 Turbo
Premium model access	Limited to 5 GPT-4 queries per day	Unlimited GPT-4, Claude 3, and others
Fallback behavior	Error shown when premium model unavailable	Automatic fallback to next available model
Rate limit reset	Every 24 hours	Every 3 hours
Regional model blocking	Full restriction in unsupported countries	Partial restriction — some models still work

You can now configure your Perplexity settings to avoid the “Model Not Available” error by selecting a stable default model and checking your subscription plan. If the error persists, use the mobile app or switch to a different model manually. For advanced users, create a browser bookmark that appends ?model=perplexity-pro to the URL to force a fallback model on every search.