Integration

Codex CLI with a custom provider: an agent without a token bill

Codex is great in a loop: tests, edits, tests again. With per-token billing every iteration costs more. Add Tenzo Cloud as a separate provider and keep your existing settings in place.

Get a free keyfrom $11 a month per stream · 3 hours with every model, no card

Why connect it

Codex CLI supports third-party providers in ~/.codex/config.toml. Tenzo Cloud is added as its own provider, so switching back takes one line.

The key lives in an environment variable rather than the file, so it will not end up in a commit by accident.

How to connect Codex CLI

You need a key from the dashboard: 3 free hours start right after sign-up. Replace tc-your-key with your key.

  1. Get a key

    Create a key in the dashboard under API keys.

  2. Store the key in a variable

    Add TENZO_API_KEY to your environment or to ~/.zshrc / ~/.bashrc.

    terminal
    export TENZO_API_KEY=tc-your-key
  3. Add the provider

    Open ~/.codex/config.toml (create it if needed) and add the block below.

    ~/.codex/config.toml
    model = "glm-5.3"
    model_provider = "tenzo"
    
    [model_providers.tenzo]
    name = "Tenzo Cloud"
    base_url = "https://tenzo-cloud.com/v1"
    env_key = "TENZO_API_KEY"
    wire_api = "responses"
  4. Run codex

    Codex now works through Tenzo Cloud with the chosen model.

Common errors and fixes

Codex complains the key is missing
TENZO_API_KEY is not set in the current terminal. Run export and restart codex.
401 invalid_api_key, “key not found”
Make sure the whole key is copied with no spaces or quotes and is enabled in the dashboard. Expired keys also return 401.
403 no_access, “no access to this model”
There is no active subscription for this model, or the key does not allow this model or your IP. Check Plans and the key restrictions in the dashboard.
404 model_not_found
The model id is wrong. Use an id from the model list, for example qwen-3.8-max or glm-5.3.
429 concurrency_limit
All your streams are busy or you exceeded 120 requests a minute. Retry in a few seconds or buy more streams, they stack.
503 model_unavailable
The model is under maintenance. Each model’s status is on the status page; switch models with one line.

Pricing

One subscription covers this tool, the chat and your own scripts.

  • Qwen 3.8 Max

    Operational
    1 stream4+ streams, each
    3 days$2.73₸1,150$2.10₸880
    7 days$5.33₸2,280$4.10₸1,750
    30 days$14.30₸6,180$11₸4,750
  • GLM 5.3

    Coming soon
    1 stream4+ streams, each
    3 days$2.86₸1,240$2.20₸950
    7 days$5.72₸2,470$4.40₸1,900
    30 days$15.73₸6,830$12.10₸5,250

Price per stream. Volume discount: 2–3 streams pay 88%, 4 or more pay 77% of the single-stream price. Streams and periods stack. Prices sync from the dashboard automatically.

FAQ

Can I keep several providers?

Yes. The provider is chosen with model_provider; other provider blocks can stay in the file.

How much does it cost and is there a token limit?

A subscription to a model at a fixed price per stream for 3, 7 or 30 days. Tokens are not counted; only concurrent requests and 120 requests a minute per account are limited. The first 3 hours are free, no card.

Do I need a VPN?

No. The API, site and chat work directly from Kazakhstan, Russia and the CIS.

Are my requests stored?

No. Prompts and answers sent through the API are not saved, neither in the database nor in logs. The journal keeps only time, model, status and token count.

Try it on your own work

Three free hours with every model, no card and no VPN. Connect Cursor, Claude Code or your own code in a minute.