--- title: What should I do if I receive a 503 error on Vercel? description: Learn about when Serverless Functions return a 503 status code and what can be done about them. url: "https://vercel.com/kb/guide/what-should-i-do-if-i-receive-a-503-error-on-vercel" published: 2025-11-03 last_updated: 2026-08-24 authors: Holden Altaffer install_vercel_plugin: npx plugins add vercel/vercel-plugin --- A **503: FUNCTION\_THROTTLED** error on your [Vercel Serverless Function](https://vercel.com/docs/concepts/functions/serverless-functions) indicates that the default limit of **concurrent executions** has been reached (30,000 for Pro and 100,000 for Enterprise). Although rare, this guide discusses potential causes and steps to resolve or mitigate this issue. ## Understanding Concurrency Concurrency refers to the ability of your Serverless Function to handle multiple executions simultaneously. ## Identifying Concurrent Issues Exceeding the concurrency limit may indicate a deeper issue. If the serverless execution time per request is increasing, it suggests that the available function capacity is being utilized fully. It's advisable to check your upstream provider (API, Content Management System, Database, etc.) to ensure it scales alongside Vercel, as increasing concurrency may strain your upstream provider further. ## Vercel Function Autoscaling Vercel automatically scales the number of function instances based on demand, up to a maximum of **30,000** concurrent executions for Pro and **100,000** for Enterprise with the option for extended concurrency. This ensures optimal performance during traffic surges without requiring manual intervention. Learn more about [automatic concurrency scaling](https://vercel.com/docs/functions/serverless-functions/runtimes#automatic-concurrency-scaling) for Vercel Functions. ## Burst Concurrency Burst concurrency allows for a temporary increase in the concurrency limit to handle sudden traffic influxes. After the initial burst, Vercel continues to scale your function by a maximum of **500 concurrency per minute**. The scaling process might take several minutes, especially during substantial traffic increases, to ensure a smooth scaling experience that aligns with natural traffic patterns and minimizes impact on performance. Learn more about [burst concurrency](https://vercel.com/docs/functions/serverless-functions/runtimes#automatic-concurrency-scaling) on Vercel. ## Other Solutions ### Caching Headers If your response can be cached, even for a short duration, adding appropriate `cache-control` headers can help reduce the number of requests reaching your function, while also improving response times. Learn more about [caching on Vercel.](https://vercel.com/docs/edge-network/caching) ### Incremental Static Regeneration (ISR) Utilize [Incremental Static Regeneration (ISR)](https://vercel.com/docs/incremental-static-regeneration) to generate static pages on-demand, which can help manage high-traffic events effectively. ### Limiting Function Maximum Duration You can limit the [maximum duration for your functions](https://vercel.com/docs/functions/configuring-functions/duration) to ensure they do not run longer than necessary, which can also help in managing resources better. ## Related Documentation - [Vercel Functions](/docs/functions) ## More Vercel Functions guides - [Optimize Runtime Cache on Vercel](/kb/guide/optimize-runtime-cache-reads-on-vercel): Diagnose rising Runtime Cache usage, trace reads to tags and code paths, and reduce avoidable reads while checking write cost and origin traffic. - [How to stop Vercel Functions from timing out](/kb/guide/what-can-i-do-about-vercel-serverless-functions-timing-out): Vercel Functions that time out usually trace back to a few causes. Learn how Fluid Compute fixes most of them and how to handle the edge cases. - [The Complete Guide to Vercel Connect](/kb/guide/vercel-connect): Use Vercel Connect to call provider APIs like Slack, GitHub, Linear, Microsoft, Discord, Snowflake, and Salesforce from your agents and services with short-lived, scoped tokens instead of long-lived secrets. ## Reference Architecture Choose a stack and work through the decisions behind it, with curated documentation, guides, templates, and courses to help you build. - [Agent stack](/kb/architecture/agents): Build and operate agents with models, tools, and durable workflows. - [Web applications](/kb/architecture/web): Choose rendering, data, authentication, and caching approaches. - [Backends](/kb/architecture/backends): Choose a runtime, connect services, and run background work. [Browse all stacks](/kb/architecture) ## Build with prompts Build an app, extend an agent, or improve your project with ready-to-use prompts for your coding agent. [Browse the prompt library](/kb/prompts) ยท [Read the Markdown index](/kb/prompts.md)