Boat.dev vs Freestyle

Boat.dev is cheaper. Freestyle builds stable, consistent Linux VMs for businesses, with memory snapshots, advanced networking, and credentials held outside the sandbox.

How I built Blendmux, Cursor for Blender

Blender has never had its Cursor moment. Here's how I built one.

Sprites Connectors is a great example of Terrible API Design

What Sprites Connectors gets wrong, how to avoid the same mistakes, and how to design APIs for agents.

Dedicated vs shared LLM inference: when reserved capacity makes sense

A decision framework for teams with steady production traffic: what shared, token-priced inference gives you, what dedicated capacity changes, and the signals that tell you it is time to switch.

How to benchmark LLM inference before you commit to capacity

What to measure when evaluating an inference endpoint for production: time to first token, per-token latency, throughput under concurrency, tail behavior, and how to make the test match your real traffic.

How to size dedicated LLM inference for steady production traffic

A capacity planning walkthrough for reserved inference: measuring tokens per second, separating prompt from completion load, choosing headroom, and comparing the result against per-token pricing.

LLM inference for coding agents: why steady, long-context traffic favors dedicated capacity

Coding agents produce a distinctive inference load: long prompts, many sequential calls per task, and traffic around the clock. Here is what that means for latency, throughput, and whether to reserve capacity.

Moving chat, extraction, and multi-step workflows to dedicated inference with an OpenAI-compatible API

A migration guide for production workloads: what OpenAI compatibility does and does not cover, how to cut over safely, and what changes for high-volume chat, document extraction, and scheduled multi-step pipelines.

A review of the adaptability of sandboxes

How Freestyle, E2B, Blaxel, Daytona, Fly Sprites, Modal, exe.dev, and Boat.dev compare on machine size, runtime type, networking, and snapshots.

Your agent is waiting for approval. What happens to its computer?

A six-provider comparison of what sandbox pause, standby, stop, and resume actually preserve while an AI agent waits for a human.

Sandbox networking APIs: Freestyle, E2B, Daytona, Blaxel, and Box

A practical, documentation-based comparison of sandbox ingress, egress controls, private networking, custom domains, and raw protocol access.

Run AI-Generated Python Without Production Credentials

Use Freestyle network secret injection to let AI-generated Python call APIs and query Postgres without putting API keys or database passwords inside the VM.

AI Security Agents Need a Lab They Can Break

A useful AI security agent reproduces findings and verifies fixes. Give it a disposable Linux lab with private networking, realistic services, and recoverable state.

16 questions you should ask before choosing a sandbox provider

A practical checklist for comparing sandbox providers across capabilities, lifecycle, access, security, networking, and scale.

Designing APIs for Agents

Good design for agents is not the same as good design for humans.

What Is a Daytona Snapshot and How It Compares

Daytona snapshots are Docker/OCI-based sandbox templates. Freestyle snapshots are full VM snapshots with memory, disk, and CPU state. That difference matters for AI agents.

The Top 6 Most Powerful Sandboxes for AI Agents

A ranked comparison of the strongest sandbox platforms for AI agents, app builders, code interpreters, browser agents, and heavy production workloads.

How to Build a Code Interpreter on Freestyle

A practical guide to building a code interpreter for AI agents on Freestyle VMs: sandboxed code execution, snapshotting, live forking, and suspend/resume sessions.

How to Build Deep Research Agents on Freestyle

A practical guide to building deep research agents on Freestyle VMs: snapshot findings, live-fork agent state, and run parallel exploration without redoing work.

How to Deploy Long-Running Agents on Freestyle

A practical guide to running long-running agents on Freestyle VMs: persistent sessions, suspend and resume, snapshots, live forking, and stateful microVMs for AI agents.

How to Do Reinforcement Learning on Freestyle

How to run reinforcement learning for AI agents on Freestyle VMs: cache a golden environment, fan out parallel rollouts, collect trajectories, train at scale.

How to Run AI Evaluations on Freestyle

A practical guide to running AI evaluations and agent evals on Freestyle VMs: ephemeral microVMs per test case, full traces, and burst parallelism.

Git Agents

Git agents are AI agents that use Git as their workspace, memory, and review layer. Here is how Git agents should branch, commit, diff, rollback, and ship with Freestyle Git.

Version Control for AI Agents

Version control for AI agents means branchable, reviewable, auditable state for AI-generated code, docs, assets, and workflows. Freestyle Git gives agent platforms real Git repos through an API.

8 Things You Didn't Know Git Already Did

Git is not just commit, push, and pull. It already ships with debugging, parallel workspaces, rollback, patch review, and agent-friendly history tools.

The best filesystem for AI agents is Git

The best way to handle an agent's filesystem is Git: normal files for the agent, and commits, branches, diffs, rollback, permissions, and review for your product.

Deprecating the Dev Server API

Announcing the deprecation of our Dev Server API and the transition to VMs

Our teams' favorite Rust crates 2025

A list of our favorite Rust crates that we think all Rustaceans should know about.

The Future of AI Coding Agents

Where AI App Builders and AI Software Engineers are headed next.

The missing trust model in AI Tools

The tools we give AI are not safe, that's going to cause problems.

How we revamped our Docs for AI

What do docs look like if people don't read them?