Overview
Veyyon is a terminal coding agent. It is installed with a one-line installer on macOS, Linux, or Windows, and then run as veyyon or vey in a repository. Authentication can be done with /login <provider> or by setting a provider API key environment variable, and local Ollama, LM Studio, or llama.cpp discovery is available when the daemon is running. The CLI, TUI, providers, and session loop are TypeScript on Bun, while grep, PTY/shell, hashline apply, and related hot paths use @veyyon/natives. The agent session tool loop follows model stream → tools (read, bash, edit, …) → results → next turn, with plan/goal/vibe modes changing tools and prompts in the engine. Model selection supports catalog providers such as Anthropic, OpenAI, Google, Groq, and OpenRouter, as well as local discovery. Roles such as smol, slow, plan, task, and advisor are configured via modelRoles. The binary includes agent orchestration, unified structural search, autonomous deep review, kernel resource limits, and deterministic line-anchored editing. It also provides configurable permission tiers, non-destructive session branching, and broad extensibility via MCP servers, custom skills, and native tools.
Pricing
Free
Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.
Visit WebsitePrices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.
Use Cases
- Teams use Veyyon to install and start a terminal coding agent in a repository.
- Teams use Veyyon to select models and roles for different tasks.
- Teams use Veyyon to inspect and manage concurrent agents across a session tree.
- Teams use Veyyon to search code with text, file, and structural AST modes.
- Teams use Veyyon to automatically review complex multi-file changes after edits.
- Teams use Veyyon to limit CPU usage across the entire session process tree.
- Teams use Veyyon to reference vault-stored secrets without exposing values in prompts.
- Teams use Veyyon to run autoresearch with multiple candidate arms and cross-review.
Features
Run a local terminal coding agent
Make edits with hashline references
Use per-project Argot shorthand
Control actions with approval modes
Choose from multiple AI providers
Work through CLI or TUI
Reviews
No reviews yet. Be the first to share your experience!