# getkimchi/kimchi

Terminal coding agent powered by Kimchi's multi-model orchestration

Repository: https://github.com/getkimchi/kimchi
Canonical: https://ross.abutalabs.com/products/getkimchi-kimchi
Homepage: https://kimchi.dev
Language: TypeScript
License: Apache-2.0
License Family: permissive
Last push: 2026-08-26T19:10:48+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 10
- inputs: {"age_days": 148, "days_push": 7, "days_rel": 7, "gap_med": 0, "n_releases_24m": 200}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2214, forks 132 (observed 2026-08-28T04:06:27.045378+00:00)

## What it is
Kimchi is a terminal-based AI coding agent CLI that autonomously executes multi-step development tasks like refactoring, writing tests, and opening PRs. It uses multi-model orchestration to route each subtask (planning, building, reviewing) to the most suitable LLM, reducing cost and avoiding vendor lock-in.

## Use cases
- implement a feature and open a PR from a single prompt
- refactor a module to async/await automatically
- write tests for a service without babysitting the agent
- route coding tasks to cheaper open-weight models to cut LLM costs
- run an autonomous plan-approve-execute workflow in the terminal
- switch between frontier models and multi-model mode mid-session

## When to choose
- you want an autonomous terminal coding agent similar to Claude Code but with model routing
- you want to use open-source or open-weight models instead of proprietary-only providers
- you want hands-off milestone-based workflows that end in a PR
- you want pay-per-token pricing with smart cost-cutting routing

## When to avoid
- you prefer inline IDE code completion like Copilot-style suggestions
- you need a GUI-based coding assistant
- you require strict single-vendor model support or air-gapped operation without their infrastructure
- you want a fully self-hosted agent without any Kimchi service dependency

## Facets
- artifact type: cli-tool
- maturity: active
- function: agent-framework, llm-inference, cli, developer-tools, prompt-engineering
- domain: developer-tools, large-language-models
- platform: windows, cli
- tags: coding-agent, multi-model-orchestration, terminal-ui, ai-pair-programming, autonomous-agent, llm-routing, ai-agents, command-line, macos, linux, nodejs

## Member repositories
- getkimchi/kimchi (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:27.045378+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:45:59.651282+00:00, confidence not recorded.
  - readme: https://github.com/getkimchi/kimchi (fetched 2026-08-28T04:06:27.045378+00:00, sha e8aee7bf3889)
  - homepage: https://kimchi.dev (fetched 2026-08-29T10:26:32.706979+00:00, sha 918e29a9e551)
  - site_page: https://docs.kimchi.dev/ (fetched 2026-08-29T10:26:32.711496+00:00, sha 2e6d9bc848b5)
  - site_page: https://kimchi.dev/pricing (fetched 2026-08-29T10:26:32.709769+00:00, sha fe47512be546)
- Data as of 2026-08-30T08:39:29.467469+00:00.