Jatevo

Jatevo exposes an OpenAI-compatible chat completions surface at https://2.lb.jatevo.ai/v1. In Anvia, configure OpenAIClient with that baseUrl, then pass Jatevo model ids to completionModel(...). The gateway also exposes a /v1/responses endpoint for stateful sessions and a /backend-api/codex endpoint for Codex CLI compatibility.

Create the Client

import { AgentBuilder } from "@anvia/core";
import { OpenAIClient } from "@anvia/openai";

const client = new OpenAIClient({
  baseUrl: "https://2.lb.jatevo.ai/v1",
  apiKey: process.env.JATEVO_API_KEY,
});

const model = client.completionModel("auto");

const agent = new AgentBuilder("support", model)
  .instructions("Answer support questions clearly.")
  .build();

const response = await agent.prompt("Hello!").send();

console.log(response.output);

baseUrl makes Anvia use the OpenAI-compatible chat completion adapter. The model id is the Jatevo id, not an Anvia-specific alias. Pass "auto" to let Jatevo route the request across its available models.

Get the Model List

Jatevo exposes a /v1/models endpoint that returns the model ids available to your key. Because the client was created with baseUrl, listModels() calls Jatevo's /models endpoint. The list call is free and does not consume quota.

const models = await client.listModels();

console.table(
  models.data.map((model) => ({
    id: model.id,
    name: model.name,
    contextLength: model.contextLength,
  })),
);

Use the id field directly with completionModel(...).

Available Models

Jatevo aggregates models from DeepSeek, Z.AI, NVIDIA, Moonshot, Alibaba, and Cerebras. Sample ids and pricing per 1M tokens:

Prices shown above were captured on June 15, 2026. Discounts, regional pricing, and available models change frequently, so always confirm the latest rates on jatevo.ai before quoting or budgeting.

The auto model id lets Jatevo's load balancer pick the best available route for the request. Some model routes are request-access only and may not be available to every key.

Notes

• Jatevo API keys use the sk-clb-… prefix and are passed as bearer tokens in the Authorization header.
• Jatevo enforces a daily request quota that resets at 00:00 UTC. Quota size scales with the wallet's $JTVO holdings.
• The gateway host is currently 2.lb.jatevo.ai. If it moves to jatevo.ai, only the baseUrl host needs to change.
• A Codex CLI compatibility endpoint is exposed at /backend-api/codex for tools that expect that path.
• Errors follow the standard OpenAI shape. Common status codes: 401 (invalid or missing key), 403 (key disabled), 429 (daily quota exhausted), 502/504 (upstream route unavailable, retry while Jatevo reroutes).
• Model capabilities still depend on the upstream model. Test the specific model id for tool calling, structured output, streaming, and multimodal support before enabling those features.

For current Jatevo API details, see the Jatevo documentation and the Jatevo model catalog.