NeverMetered
Get API key

Use NeverMetered with Hermes Agent

Nous Research's open-source terminal agent, using a custom OpenAI-compatible endpoint.

OpenAI Chat
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 “Hermes Agent”. 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 Hermes Agent

    Already have Hermes Agent? 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. Then open a new terminal window so the hermes command is found.

    curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
  3. 3

    Point Hermes at the gateway

    Run these five commands. They set the endpoint and model, and keep your key in Hermes' own ~/.hermes/.env (the config file refers to it by name), so there are no prompts to answer.

    hermes config set model.provider custom
hermes config set model.base_url https://api.nevermetered.com/v1
hermes config set model.default qwen3.8-27b
hermes config set NEVERMETERED_API_KEY YOUR_API_KEY
hermes config set model.api_key '${NEVERMETERED_API_KEY}'
  4. 4

    Or use Hermes' setup wizard

    Prefer prompts? Run this and pick Custom endpoint (enter URL manually), near the bottom of the provider list (the wording changes between Hermes versions). Enter https://api.nevermetered.com/v1 as the API base URL and YOUR_API_KEY as the key. Then press Enter to keep the default compatibility mode, press Enter to accept the detected model (qwen3.8-27b), leave the context length blank, and accept the display name. The last screen picks a reasoning effort; Skip is fine.

    hermes model
  5. 5

    Verify it works

    Run a one-off prompt, or start hermes and say hello. The request shows up in Requests within a few seconds, with its input, cached and output tokens.

    hermes chat -q "Reply with the word ready" --oneshot
Tips
  • Hermes needs a model with at least 64K tokens of context and won't start below that. Your plan's window is 128K, which is enough.
  • The base URL must end in /v1. Hermes reads the model and its context window from the gateway's model list, so there is no model name to type in the wizard.
  • Each new Hermes session also makes one small background request to title the session, so on the free plan starting a session uses two of your daily requests.
  • Hermes sends a reasoning level with every request (medium by default). Change it for one run with hermes chat --reasoning <level>, or for every run with hermes config set agent.reasoning_effort <level>.

Troubleshooting

hermes: command not found
Open a new terminal window: the installer adds Hermes to your path for terminals started afterwards. If it still isn't found, run source ~/.zshrc (or source ~/.bashrc).
Model … has a context window of … which is below the minimum 64,000
Hermes reads the window from the gateway and needs at least 64K. Check your plan's window on Plan & billing, or contact support.
Model '…' isn't available on Custom endpoint
Use the model id exactly as the gateway lists it: hermes config set model.default qwen3.8-27b.
Hermes seems frozen and never replies
On the free plan, once the daily requests are used up Hermes can wait for the allowance to reset instead of showing an error. Check Usage, or upgrade for no daily limit.
The wizard has no "Custom endpoint" entry
The wizard's wording changes between Hermes versions. The five commands above don't use the menu.
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.