Skip to main content

Create a key

Create one in the developer console at app.trynito.ai/api, then export it:

Send a request

With curl, or with the OpenAI SDK pointed at Nito (pip install openai or npm install openai). The base URL ends in /v1; the SDK appends /chat/completions.
The reply is at choices[0].message.content; usage reports the tokens used.
That is it. You are calling Nito.

Try Next

  • Switch models. Change model and nothing else: moonshotai/kimi-k3 (open-weight, Private), anthropic/claude-sonnet-5 or openai/gpt-5.5 (frontier, Anonymous). GET /v1/models is the live list. See Model List.
  • Run inside attested hardware. On models with an attested build, the :confidential variant, for example z-ai/glm-5.2:confidential, runs inside sealed hardware and returns attestation evidence you can verify. A model that cannot serve the level is rejected, never downgraded. See Privacy Levels.
  • Stream the reply. Add stream: true. See Stream Your First Response.

If Something Fails

  • 401: the key is missing, mistyped, or not sent as Authorization: Bearer <key>. Keys start with z_sk_.
  • 400 naming the model: the ID is not in the catalog. Copy an exact one from GET /v1/models.
  • 429 with insufficient_quota: the organization has no funding. An API key needs a subscription or a prepaid balance; add funds on Billing. See Plans.

Next

Make Your First Request

The same call, field by field.

Stream Your First Response

Tokens as they arrive.

OpenAI SDK Drop-in

Existing OpenAI code.

Privacy Overview

Levels and attestation.