Learn AI

Sign In

by Vercel

VercelFollow

How a libheif RCE Was Found, Fixed, and Disclosed

Reproducing, disclosing, and fixing the libheif vulnerability with Hacktron and the maintainers
...
...
VercelFollow

Open-weight models take 56% of token volume, Astra doubles Fable 5.1 spend

Open-weight models take 56% of token volume, Astra doubles Fable 5.1 spend
...
...
VercelFollow

How Indexed Shards Cut CDN Lookup Latency by 91%

How we cut CDN metadata lookup latency by 91%
...
...
VercelFollow

Fluid Compute: How Machines Adapt to Every Workload

Compute that takes any shape
...
...
VercelFollow

How Agents Build Better On-Brand Pages with design.md

How our agents build on-brand pages with design.md
...
...
VercelFollow

How TypeScript Turns Ordinary Code into Durable Workflows

The best workflow engine is a programming language
...
...
VercelFollow

Run SDK: Safely Execute Agent-Generated Code

Introducing Run SDK: secure eval for your agents
...
...
VercelFollow

How v0 Uses Snowflake Without Exposing OAuth Tokens

How v0 authenticates to Snowflake without exposing the user's OAuth token
...
...
VercelFollow

How to Build a Secure AI Software Factory

Building a software factory for AI SDK
...
...
VercelFollow

Migrating Vercel’s Build State from Redis to DynamoDB

How we migrated the database behind every Vercel build
...
...
VercelFollow

AI Model Usage Shifts: DeepSeek Rises, Costs Fall

DeepSeek overtakes Google on volume, cost per token falls 13.6%
...
...
VercelFollow

AI Is Changing Cybersecurity: Defend Your Code Now

Everything hackable will get hacked
...
...
VercelFollow

Sandbox Security: Control What Untrusted Code Can Reach

A sandbox without a network boundary is only half a sandbox
...
...
VercelFollow

Agent Plugins: Package Skills and MCP Servers Once

Introducing Agent Plugins
...
...
VercelFollow

Next.js 16.3 on Vercel: Faster Apps, Better Insights

Vercel supports Next.js 16.3
...
...
VercelFollow

How Factory Scaled Next.js to Tens of Millions of Requests

How Factory scaled its cloud backend to tens of millions of daily requests on Vercel
...
...
VercelFollow

How AI Models Find Vulnerabilities—and What It Costs

DeepsecBench: evaluating model performance in finding cybersecurity vulnerabilities
...
...
VercelFollow

How Searchable ships customer-requested features in 30 minutes on Vercel

How Searchable ships customer-requested features in 30 minutes on Vercel
...
...
VercelFollow

How Speechify serves 500,000 dynamic pages to 60 million users on Vercel

How Speechify serves 500,000 dynamic pages to 60 million users on Vercel
...
...
VercelFollow

Open-weight models surge to 29% of volume, price per token flattens

Open-weight models surge to 29% of volume, price per token flattens
...
...