Whip Open source

What it does

Whip is a terminal coding agent written in Go: a bubbletea TUI wrapped around an LLM tool-use loop (bash, read, write, edit, and subagent tools) that keeps working on a goal until it's done. It ships as a single static binary with no runtime to install, which makes it easy to drop onto a fresh machine, a container, or a CI box without pulling in Node or Python dependencies.

What is inside

• A tool-use loop with parallel, streaming tool calls
• Background subagents and file locks so concurrent edits don't collide
• Live model discovery against provider catalogs, refreshed roughly every 24 hours
• Built-in commands: /goal <text> to run to completion, /model to switch models, /context-doctor to audit token injection, /mcp to list connected MCP servers, and a /theme picker

Works with

Whip is provider-agnostic rather than tied to one model vendor. It defaults to inference.net's open-source model catalog, but talks to any OpenAI-compatible endpoint, including OpenRouter (browse the full catalog with /model) and a ChatGPT/Codex subscription via whip auth codex. It reads MCP servers from a .mcp.json file in the repository, so an existing MCP configuration carries over without extra setup.

Pricing

Whip itself is free and open source under the Apache-2.0 licence, with no separate cost to download, build or run the binary. Your actual spend depends entirely on which backend you point it at: inference.net's open-source model catalog, a pay-as-you-go OpenRouter account, or an existing ChatGPT/Codex subscription authenticated via whip auth codex. No official Whip pricing page exists beyond the GitHub README, since the tool is only a client for those providers.

How to install or connect

curl -fsSL https://raw.githubusercontent.com/context-labs/whip/main/install.sh | sh

Or build from source with Go 1.27+:

go install github.com/context-labs/whip/cmd/whip@latest

Update an existing install with whip update.

Maintenance and safety

Active project (656+ commits, weekly star growth, pushed within the last day) under the Apache-2.0 licence. The install script fetches over HTTPS from the project's own GitHub repo, but as with any curl-pipe-to-shell installer, reviewing install.sh first is sensible before running it with elevated trust.

Who should use it

Developers who want a lightweight, dependency-free terminal agent for running open-source or self-hosted models, or an existing Codex subscription, without committing to a heavier IDE-based tool or a proprietary hosted backend. It's also a reasonable pick for CI or container environments where a single static binary is easier to provision than a Node- or Python-based agent.