NeverMetered
Get API key

Use NeverMetered with Codex CLI

OpenAI's open-source terminal agent, using a custom Responses API provider.

OpenAI Responses
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 “Codex CLI”. 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 Codex CLI

    Already have Codex CLI? 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.

    curl -fsSL https://chatgpt.com/codex/install.sh | sh
  3. 3

    Add a model provider

    Add this to your Codex config (create the file if it doesn't exist). Codex reads the key from the environment variable named in env_key.

    ~/.codex/config.toml
    model = "qwen3.8-27b"
model_provider = "nevermetered"
model_context_window = 131072

[model_providers.nevermetered]
name = "NeverMetered"
base_url = "https://api.nevermetered.com/v1"
env_key = "NEVERMETERED_API_KEY"
wire_api = "responses"
  4. 4

    Export your key and start Codex

    export NEVERMETERED_API_KEY="YOUR_API_KEY"
codex
  5. 5

    Keep it for new terminals (optional)

    The command above only lasts for the current terminal. Save it once so every new terminal has it.

    # zsh (the macOS default)
echo 'export NEVERMETERED_API_KEY="YOUR_API_KEY"' >> ~/.zshrc
# bash
echo 'export NEVERMETERED_API_KEY="YOUR_API_KEY"' >> ~/.bashrc
  6. 6

    Verify it works

    Run a non-interactive task. The request shows up in Requests within a few seconds, with its input, cached and output tokens.

    codex exec "Reply with the word ready"
Tips
  • model_context_window tells Codex your plan's 131,072-token window, so it can tell when a conversation is getting close. To make it compact earlier, set model_auto_compact_token_limit (a token count) next to it.
  • Codex resends the conversation on every turn. Requests from one session stick to the same engine for a while, so the shared prefix is served from cache and shows up as cached input tokens.
  • Codex can run several tool calls and agents at once; each generation uses one of your parallel streams. Your plan allows up to 65,536 output tokens per request.

Troubleshooting

Codex says the environment variable is missing
Export NEVERMETERED_API_KEY in the shell that runs codex, or add it to your shell profile.
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.