무제한 Claude에서 Kilo Code 실행

Kilo Code는 전체 파일 컨텍스트로 에이전트 루프를 돌려 토큰 처리량이 높습니다. 정액 무제한이면 처리량이 비용 문제가 아닙니다.

Pick the provider

In Kilo's settings choose the API Provider. For the gateway's OpenAI-compatible surface, select OpenAI Compatible, set the Base URL with /v1, paste your unlimited key, and type the model id. For native Anthropic, choose Anthropic with Use custom base URL and the host root.

Anthropic, Claude Code, and OpenAI-Compatible are distinct provider flows — match the one to your gateway's wire format.

Model id and context

The Model ID in OpenAI Compatible mode is free-text and unvalidated, so a typo 404s as 'model not found' — enter exactly what the gateway serves. Set Context Window and Max Output Tokens under Model Configuration or Kilo may truncate.

Enable Computer Use / function-calling if the gateway forwards tool calls; agentic edits depend on it.

Why agentic coding wants unlimited

Searches like 'kilo code claude agents' and 'kilo code claude free' show users wanting to run Kilo hard without watching a meter. Each plan/edit loop fans out into many calls, and a big task can run for a long time.

Flat-rate unlimited lets you keep the strongest model selected and let the agent iterate to completion, for one fixed price.

Model selection

Use a Sonnet-class model for implementation and an Opus-class one for hard planning or debugging. On unlimited you can leave the stronger model on without cost anxiety.

One unlimited key covers every Kilo session for the plan's duration.

# Kilo Code -> Settings -> API Provider
# OpenAI Compatible:
#   Base URL: https://claudeapikey.dev/v1
#   Model ID: claude-sonnet-4-5

FAQ

Can I run Kilo Code on unlimited Claude?
Yes — set OpenAI Compatible with Base URL {SITE}/v1 (or Anthropic + custom base URL), paste your unlimited key, and the flat-rate plan covers every session.

Why does Kilo say 'model not found'?
The Model ID is free-text and unvalidated in OpenAI Compatible mode — a typo 404s. Enter the exact gateway id.

Why unlimited for Kilo Code?
Its agentic loops keep token throughput high. Flat-rate unlimited lets the agent run freely instead of rationing tokens.

Do I set the context window?
Yes — set Context Window and Max Output Tokens manually or Kilo may truncate.