NeverMetered
Get API key

Use NeverMetered with Claude Code

Anthropic's terminal coding agent, pointed at your own base URL and model.

Anthropic Messages
Model
qwen3.8-27b
128K context on the free plan65,536 max output tokens on the free planReplace YOUR_API_KEY with your key.
  1. 1

    Create an API key

    Create a key in API keys and name it after the tool, for example “Claude Code”. A dedicated key per tool lets you see its usage separately and give it its own parallel-stream cap. The key is shown once; the snippets below use YOUR_API_KEY where it goes.

  2. 2

    Install Claude Code

    Already have Claude Code? Skip this step. Open a terminal first: on macOS press ⌘ Space, type Terminal and press Return; on Windows open the Start menu, type PowerShell and press Enter. Check it worked with claude --version.

    curl -fsSL https://claude.ai/install.sh | bash
  3. 3

    Add a nevermetered-claude command

    Adds a command, nevermetered-claude, that starts Claude Code against the gateway, then runs it. Plain claude is untouched and keeps using whatever it used before, so your normal Claude Code isn't redirected: use this one for the work you want to send here. Every model variable points at the same model: Claude Code asks for its own claude-* model names for the main loop, subagents and background tasks, and those names don't exist here. ANTHROPIC_AUTH_TOKEN sends your key as a bearer token without the interactive API-key prompt that ANTHROPIC_API_KEY triggers. Your key is saved in a file in your home folder that only you can read. Open a new terminal after pasting, then run the command any time.

    cat > ~/.nevermetered-claude.sh <<'EOF'
nevermetered-claude() {
  ANTHROPIC_BASE_URL="https://api.nevermetered.com" \
  ANTHROPIC_AUTH_TOKEN="YOUR_API_KEY" \
  ANTHROPIC_MODEL="qwen3.8-27b" \
  ANTHROPIC_DEFAULT_OPUS_MODEL="qwen3.8-27b" \
  ANTHROPIC_DEFAULT_SONNET_MODEL="qwen3.8-27b" \
  ANTHROPIC_DEFAULT_HAIKU_MODEL="qwen3.8-27b" \
  CLAUDE_CODE_SUBAGENT_MODEL="qwen3.8-27b" \
  API_TIMEOUT_MS="600000" \
  CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC="1" \
  CLAUDE_CODE_MAX_OUTPUT_TOKENS="65536" \
  CLAUDE_CODE_MAX_CONTEXT_TOKENS="131072" \
  claude "$@"
}
EOF
chmod 600 ~/.nevermetered-claude.sh
case "$SHELL" in *zsh) rc=~/.zshrc ;; *) rc=~/.bashrc ;; esac
grep -qsF '.nevermetered-claude.sh' "$rc" || echo '. ~/.nevermetered-claude.sh' >> "$rc"
. ~/.nevermetered-claude.sh
nevermetered-claude
  4. 4

    Only for this terminal window (optional)

    Sets the same variables for the current terminal only and starts claude with them. Nothing is saved: a new terminal goes back to normal, and claude in this one is redirected until you close it.

    export ANTHROPIC_BASE_URL="https://api.nevermetered.com"
export ANTHROPIC_AUTH_TOKEN="YOUR_API_KEY"
export ANTHROPIC_MODEL="qwen3.8-27b"
export ANTHROPIC_DEFAULT_OPUS_MODEL="qwen3.8-27b"
export ANTHROPIC_DEFAULT_SONNET_MODEL="qwen3.8-27b"
export ANTHROPIC_DEFAULT_HAIKU_MODEL="qwen3.8-27b"
export CLAUDE_CODE_SUBAGENT_MODEL="qwen3.8-27b"
export API_TIMEOUT_MS="600000"
export CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC="1"
export CLAUDE_CODE_MAX_OUTPUT_TOKENS="65536"
export CLAUDE_CODE_MAX_CONTEXT_TOKENS="131072"
claude
  5. 5

    Or change every Claude Code session (optional)

    This sends every Claude Code session on this machine here, including any you want to keep on your Anthropic account, so most people want the command above. To do it anyway, put the same variables in the env block of your user settings, merging them into the file if it already exists.

    ~/.claude/settings.json
    {
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.nevermetered.com",
    "ANTHROPIC_AUTH_TOKEN": "YOUR_API_KEY",
    "ANTHROPIC_MODEL": "qwen3.8-27b",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "qwen3.8-27b",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "qwen3.8-27b",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "qwen3.8-27b",
    "CLAUDE_CODE_SUBAGENT_MODEL": "qwen3.8-27b",
    "API_TIMEOUT_MS": "600000",
    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "1",
    "CLAUDE_CODE_MAX_OUTPUT_TOKENS": "65536",
    "CLAUDE_CODE_MAX_CONTEXT_TOKENS": "131072"
  }
}
  6. 6

    Verify it works

    Run a one-off prompt with the new command. Inside an interactive session, /status shows the base URL and model in use. The request shows up in Requests within a few seconds, with its input, cached and output tokens.

    nevermetered-claude -p "Reply with the word ready"
Tips
  • CLAUDE_CODE_MAX_CONTEXT_TOKENS tells Claude Code your plan's 131,072-token window. It doesn't recognise this model's name, so without it Claude Code assumes 200K and compacts the conversation too late. It may still print an "isn't described by this version's model catalog" note; that's expected.
  • Each subagent Claude Code starts runs as its own stream, so parallel subagents count against your parallel-stream limit. Extra requests wait in your queue allowance.
  • Claude Code makes small background requests (for example session titles). On the free tier those count toward your daily requests; CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=1 cuts the optional ones.
  • Requests can wait in the fair queue when the service is busy, so keep API_TIMEOUT_MS generous. Your plan allows up to 65,536 output tokens per request.
  • Unset ANTHROPIC_API_KEY in the same shell so Claude Code doesn't warn about two competing credentials.

Troubleshooting

model_not_found naming a claude-… model
One of the model variables above is missing, so Claude Code fell back to its own model name. Set all of them, then restart claude.
Claude Code still talks to Anthropic
You started plain claude, which is deliberately left alone. Run nevermetered-claude instead (open a new terminal first if the command isn't found) and check /status. If you used the terminal-only variables, they only exist in the window you pasted them into.
404 model_not_found
Use a model id exactly as GET /v1/models lists it, for example qwen3.8-27b. Plans can restrict which models a key may call.
400 context_length_exceeded
The conversation outgrew the 128K context window (input + output). Let the tool compact or start a fresh session, or get a larger window on your plan.
429 concurrency_limit_exceeded
All your parallel streams are busy and your queue allowance is full. Lower the tool's parallelism (subagents, batch size), or get more streams on your plan.
Answers stop mid-sentence
The response hit your plan's output cap of 65,536 tokens (finish reason length or max_tokens). Ask for shorter steps, or upgrade for a higher cap.
403 email_verification_required
Free requests need a verified email. Click the link in your verification email (you can resend it from the dashboard), or upgrade.
429 daily_quota_exceeded
You've used today's free requests; agents send one request per step, so they go quickly. The allowance resets at 00:00 UTC, and paid plans have no daily limit.
503 / 504 request_queue_timeout
The service stayed busy longer than the queue timeout and nothing was generated. Retry with backoff, and give the tool a generous request timeout.
503 service_unavailable
No GPU capacity is online at the moment. It's temporary and your setup is fine: retry in a little while, and check the status page.