GeezerVison is showing www.fastly.com — Edge Computing & Serverless Platform | Fastly

Deploy

Code locally, deploy globally. With absolute safety

The fastest, most secure execution environment for your mission-critical workloads.

Code, deploy, and scale - milliseconds from your users

Everything your engineering teams need to write, deploy, and scale code closer to your users - from compute and data stores to MCP servers and dev tooling.

Edge Compute

Fastly Compute is a globally distributed development platform providing the fastest, most secure, and scalable serverless environment to run your code at the edge.

Object Storage

S3-compatible object storage with zero egress fees. Migrate easily with Direct Fetch to store images, video, and AI datasets directly at the edge, without breaking your budget.

Key Value Store

Fastly KV Store simplifies edge data management by delivering a high-speed, globally distributed storage platform, including support for secure key-value storage and scalable configuration management at the edge

Websockets & Fanout

Deliver real-time WebSockets and pub/sub messaging at global scale. Fanout handles millions of connections at the edge so your origin doesn't have to.

AI Accelerator

Speed up LLM apps with semantic caching that recognizes similar queries and serves cached answers instantly, cutting latency and API costs.

MCP Server

Control Fastly through natural language. Combined with the Fastly AI Agent Toolkit, your AI coding agent gets secure control plus the skills to use it correctly.

Programmable Cache

Get full programmatic access to the same legendary caching that powers our CDN, with HTTP, Core, and Simple Cache APIs for total control.

Developer SDKs

Write Compute code in Rust, JavaScript, Go, or C++ with official SDKs, API clients, and hundreds of code examples to get you building fast.

Developer Tools

CLI, APIs,Fiddle, Terraform, GitHub Actions, and more. It’s everything your team needs to build, deploy, manage, and observe apps on Fastly.

Edge Compute

Fastly Compute is a globally distributed development platform providing the fastest, most secure, and scalable serverless environment to run your code at the edge.

Learn more

Object Storage

S3-compatible object storage with zero egress fees. Migrate easily with Direct Fetch to store images, video, and AI datasets directly at the edge, without breaking your budget.

Learn more

Key Value Store

Learn more

Websockets & Fanout

Deliver real-time WebSockets and pub/sub messaging at global scale. Fanout handles millions of connections at the edge so your origin doesn't have to.

Learn more

AI Accelerator

Speed up LLM apps with semantic caching that recognizes similar queries and serves cached answers instantly, cutting latency and API costs.

Learn more

MCP Server

Control Fastly through natural language. Combined with the Fastly AI Agent Toolkit, your AI coding agent gets secure control plus the skills to use it correctly.

Learn more

Programmable Cache

Get full programmatic access to the same legendary caching that powers our CDN, with HTTP, Core, and Simple Cache APIs for total control.

Learn more

Developer SDKs

Write Compute code in Rust, JavaScript, Go, or C++ with official SDKs, API clients, and hundreds of code examples to get you building fast.

Learn more

Developer Tools

CLI, APIs,Fiddle, Terraform, GitHub Actions, and more. It’s everything your team needs to build, deploy, manage, and observe apps on Fastly.

Learn more

Inherent memory safety

Fastly Compute runs every request in its own sandbox by default, wiping memory and state between requests. This makes it hard and expensive for attackers to compromise your apps without you having to do anything more than simply run your code on our platform.

Isolated WebAssembly sandbox

Run each workload in a highly isolated environment, with zero shared resources.

No data left behind

The platform wipes memory and state between requests to prevent data leakage.

Accelerate release cycles

Robust, built-in security simplifies deployment, reducing the need for manual security reviews.

Eliminate cold starts

Our high-performance WebAssembly runtime instantiates workloads instantaneously upon request, executing code in microseconds. Eliminate serverless cold starts to deliver a consistent, hyper-fast experience that directly improves conversion and engagement.

Instant global deploys

Push new code globally in seconds, not minutes, with zero downtime.

Improve conversions and engagement

Your code runs in microseconds, providing a consistent, low-latency experience for every user.

Instant global scale

No manual scaling or regional provisioning needed. Instant global scale is a platform feature.

Use standard languages and toolchains

Stop compromising with proprietary runtimes and vendor lock-in. Build with standard languages such as Rust, C++, Go, and JavaScript, integrating seamlessly into your existing CI/CD pipelines and toolchains. What you test locally behaves the same in production.

No proprietary runtimes

Build with standard JS, Rust, Go, and C++, or utilize one of our community supported SDKs.

Learn more

Familiar developer tooling

Integrate seamlessly with your existing CI/CD pipelines and local test environments.

Learn more

Local-to-global fidelity

The code you test locally with the CLI, behaves the same way in production.

Learn more

Slash origin costs

Slash origin costs and eliminate infrastructure bottlenecks by serving more code from the edge. Maximize ROI by running A/B testing, authentication, and personalization closer to your users, significantly reducing requests to your backend infrastructure.

Edge authentication

Validate tokens and authorize users at the edge so invalid traffic never reaches your origin.

Dynamic personalization

Run A/B tests and serve localized content at the edge without querying central databases.

Edge rate limiting

Block abusive, high-volume bot traffic at the edge before it puts strain on backend servers.

Edge geolocation

Parse location headers at the edge to serve regional content without origin roundtrips.

Agentic AI orchestration

Route and parse agentic sub-tasks closer to users, slashing costly model and origin trips.

Dynamic URL redirects

Instant edge redirects eliminate latency, preserving your crawl budget and boosting SEO.

Build personalized experiences, at scale

Spend your time focusing on your features and your users, not infrastructure. Fastly Compute gives you a secure-by-default runtime with microsecond startup, and the flexibility to run everything from personalization logic to AI agents, all on the same platform that delivers and secures your traffic.

Explore Compute

Featured use cases

1stDibsCommerce

Power seamless shopping experiences with instant page loads and dynamic images.

LayercakeMedia

Scale live video streams during massive traffic spikes while slashing operating costs.

New RelicSaaS

Utilize edge compute and observability tools as part of your multi-CDN strategy and failover systems.

Pricing built around your business

Experiment for free, pay as you grow, or work with our team on a solution tailored to your technical requirements.

Explore Pricing

Starter
For developer teams at organizations who want to get started with edge computing using familiar tools and languages with predictable pricing.
per month$500
Advantage
For developer teams at organizations who require a boost in capacity to apply edge computing to more use cases.
Contact sales
Ultimate
For developer teams at organizations embracing serverless computing, who need maximal capacity, plus additional configuration and storage options.
Contact sales

Featured resources

AIAI Agents on Fastly Compute: How It Works

Learn how to run AI agents on Fastly Compute, using the edge for low latency and Wasm sandboxes for enterprise-grade security.

ComputeBuilding a Secure MCP Server with Fastly Compute

How to build an MCP server on Fastly Compute for secure, low-latency access to your platform's API.

AIFastly Ranked as a Leader in 2026 Forrester Wave™

Fastly scores 5 out of 5 across eight criteria in the 2026 Forrester Wave™ for Edge Development Platforms.

Frequently asked questions

How quickly can I deploy code globally with Fastly Compute?

Fastly Compute lets teams deploy code globally in seconds without configuring individual regions or managing infrastructure. Workloads run across Fastly’s distributed network, close to end users, so developers can move code to production quickly while maintaining control over application versions and deployments.

How does Fastly help teams manage Compute deployments and rollbacks?

Fastly tracks Compute deployments and versions so teams can see what code is running and maintain a clear history of changes. If an issue occurs, teams can roll back to an earlier version without rebuilding or manually redeploying infrastructure.

Can Fastly Compute run AI agents at the edge?

Yes. Fastly Compute can execute AI agents and agentic sub-tasks at the edge, closer to users, APIs, and other services they interact with. Its WebAssembly-based runtime provides fast startup and workload isolation, making it well suited for latency-sensitive steps within distributed AI workflows.

Why run AI agent workloads at the edge?

Running parts of an AI agent workflow at the edge can reduce the time spent sending requests back to centralized infrastructure. Developers can move latency-sensitive tasks closer to users and services while using centralized models, data stores, or cloud infrastructure for workloads that belong elsewhere.

Do developers need to manage infrastructure to run code on Fastly?

No. Fastly manages the underlying infrastructure, scaling, and provisioning for Compute workloads. Developers can focus on writing and deploying application logic rather than selecting regions, configuring servers, or planning capacity as traffic changes.

Ready to build on Fastly?

Start free or connect with our team for a solution tailored to your infrastructure, performance, and security needs.

Start for free Talk to our team