Create chat — POST /v1/ai/mcp/chat

Asks one model one prompt and answers the reply, which model served it and who paid.

POST /v1/ai/mcp/chat

AI → · All API references → · Get started →

Address	`https://api.hanzo.ai/v1/ai/mcp/chat`
Method	`POST`
Operation	`aiChat`
Auth	`Authorization: Bearer $HANZO_API_KEY`

Asks one model one prompt and answers the reply, which model served it and who paid. The call is the caller's own POST /v1/chat/completions, run as the caller, so it is gated, billed and refused exactly as that route is; a refusal names its code and what lifts it, never an amount. An empty model asks the deployment's default Hanzo model.

Request

4 fields, body application/json (required).

Field	In	Type	Required	Description
`max_tokens`	body	integer (int64)	—	MaxTokens bounds the answer in tokens; 0 leaves it to the model.
`model`	body	string	—	Model is the model to answer, by id (zen5, enso, kai, anthropic/claude-sonnet-4.5 — aiModels lists them).
`prompt`	body	string	—	Prompt is what to ask.
`system`	body	string	—	System is the system message ahead of the prompt, when set.

Response

Status	Body	Meaning
`200`	aiChat	ok
`default`	problem-details	refused

200 body — 10 fields.

Field	In	Type	Always	Description
`answer`	body	string	—	Answer is the model's reply.
`model`	body	string	—	Model is the model asked: the caller's, or the default when it named none.
`receipt`	body	aiReceipt	—
`receipt.class`	body	string	—	Class is the model's class as billed: premium, ours or free (X-Hanzo-Usage-Class).
`receipt.fallback`	body	string	—	Fallback is the model that answered in place of the one asked, in limited mode (X-Hanzo-Fallback).
`receipt.paid_by`	body	string	—	PaidBy is plan, credits or free (X-Hanzo-Paid-By); absent when the answer named no payer, as a free model's does.
`receipt.reason`	body	string	—	Reason is the refusal code that sent the request to the fallback (X-Hanzo-Usage-Reason).
`receipt.routed`	body	string	—	Routed is the model `auto` resolved to (X-Routed-Model).
`receipt.served`	body	string	—	Served is the model that answered (X-Hanzo-Served).
`receipt.usage`	body	string	—	Usage is where that class stands for the payer: ok, near or limited (X-Hanzo-Usage).

Failure carries the platform error shape — see Errors.

Examples

CLI

SDK

HTTP

MCP

hanzo has no subcommand for this operation — the CLI serves only what cloud's live route table confirms. Use HTTP or an SDK.

.ts

.py

.go

.rs

.java

import { Configuration, AiApi } from 'hanzoai';

const api = new AiApi(new Configuration({ accessToken: process.env.HANZO_API_KEY }));
const { data } = await api.aiChat({ max_tokens: 0, model: "<model>" });

from hanzoai.cloud import ApiClient, Configuration
from hanzoai.cloud.api import AiApi

client = ApiClient(Configuration(access_token=os.environ["HANZO_API_KEY"]))
result = AiApi(client).ai_chat(max_tokens=0, model="<model>")

cfg := hanzoai.NewConfiguration()
cfg.AddDefaultHeader("Authorization", "Bearer "+os.Getenv("HANZO_API_KEY"))
client := hanzoai.NewAPIClient(cfg)

resp, _, err := client.AiAPI.AiChat(context.Background()).Execute()
if err != nil {
	return err
}

use hanzo_client::apis::{configuration::Configuration, ai_api};

let mut cfg = Configuration::new();
cfg.bearer_access_token = std::env::var("HANZO_API_KEY").ok();

let result = ai_api::ai_chat(&cfg, Default::default()).await?;

import ai.hanzo.cloud.ApiClient;
import ai.hanzo.cloud.api.AiApi;

ApiClient client = new ApiClient();
client.setBearerToken(System.getenv("HANZO_API_KEY"));

var result = new AiApi(client).aiChat();

curl -X POST https://api.hanzo.ai/v1/ai/mcp/chat \
  -H "Authorization: Bearer $HANZO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
       "max_tokens": 0,
       "model": "<model>"
     }'

MCP declares no tool for ai — tools/list on https://api.hanzo.ai/v1/mcp names the products it does reach. Use HTTP or an SDK.

AI API · All Hanzo APIs · Interactive reference

Was this page useful?

PreviousCreate agent NextCreate decisions — POST /v1/ai/mcp/decisions