Guide · Claude Code

Connect Claude Code to Synap

3 steps · about 2 minutes

  1. Get a free key

    Sign in with Google at synap.linkrra.com — no card. Your key starts with sk-pool-.

  2. Paste the setup block

    Five lines. They tell Claude Code where to send the work, and which model answers.

    export ANTHROPIC_BASE_URL=https://synap.linkrra.com
    export ANTHROPIC_AUTH_TOKEN=sk-pool-...
    export ANTHROPIC_MODEL=qwen3-coder-30b
    export ANTHROPIC_DEFAULT_HAIKU_MODEL=qwen3-coder-30b
    export CLAUDE_CODE_MAX_CONTEXT_TOKENS=98304
  3. Start it as always

    Your prompts, files and tools work exactly as before.

    claude
Notes, prices and limits

Claude Code speaks Anthropic’s Messages API, and so does Synap. Recommended model: qwen3-coder-30b — $0.0675 in and $0.27 out per 1M tokens, 98,304-token context.

The base URL has no /v1 — Claude Code appends /v1/messages itself.

ANTHROPIC_DEFAULT_HAIKU_MODEL is what Claude Code uses for its own small background tasks; pointing it at Synap keeps those on Synap too.

Without CLAUDE_CODE_MAX_CONTEXT_TOKENS Claude Code opens on a warning that it does not recognise the model. The number is this model’s context window.

Tool use and streaming behave as they do with Anthropic.

On the free tier: every request counts input and output against 100,000 free tokens a day, and answers stop at 4,096 output tokens. Coding agents send long prompts, so the free tier covers a short session; credits cover the rest.

Field names from Claude Code docs: LLM gateway configuration, checked 2026-11-01 09:50 UTC. Prices refresh from /v1/models when this page loads.