Quickstart

Make one OpenAI-compatible chat completions call with the Terminaili base URL, an API key, and ipop/coder-balanced.

OpenAI-compatible

You need three thingsBase URLhttps://api.terminaili.com/v1

API keyAuthorization headerAuthorization: Bearer $TERMINAILI_API_KEY

Model idCanonical quickstart modelipop/coder-balanced

cURL

curl https://api.terminaili.com/v1/chat/completions \
  -H "Authorization: Bearer $TERMINAILI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "ipop/coder-balanced",
    "messages": [{ "role": "user", "content": "Hello from Terminaili." }]
  }'

Python

from openai import OpenAI
import os

client = OpenAI(
    base_url=os.environ["OPENAI_BASE_URL"],
    api_key=os.environ["OPENAI_API_KEY"],
)

completion = client.chat.completions.create(
    model=os.environ.get("OPENAI_MODEL", "ipop/coder-balanced"),
    messages=[{"role": "user", "content": "Hello from Terminaili."}],
)

TypeScript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: process.env.OPENAI_BASE_URL,
  apiKey: process.env.OPENAI_API_KEY,
});

const completion = await client.chat.completions.create({
  model: process.env.OPENAI_MODEL ?? "ipop/coder-balanced",
  messages: [{ role: "user", content: "Hello from Terminaili." }],
});

Get an API key Try in playground Authentication SDK setup

Setup

Set auth and SDK environment values before adding optional API features.

Authentication

Use a Terminaili API key in the bearer header or the OpenAPI key header.

Primary headerAuthorization: Bearer $TERMINAILI_API_KEY

OpenAPI alternatex-api-key: $TERMINAILI_API_KEY

StorageKeep keys in server-side environment variables, workers, CLIs, or CI secrets.

SDK Setup

Use the normal OpenAI SDKs and override only the base URL, API key, and model.

Pythonpip install openai

Node.jsnpm install openai

EnvironmentSet the same values locally, in CI, and in deployed server code.

.env

export TERMINAILI_API_BASE_URL="https://api.terminaili.com/v1"
export TERMINAILI_API_KEY="replace-me"
export OPENAI_BASE_URL="https://api.terminaili.com/v1"
export OPENAI_API_KEY="$TERMINAILI_API_KEY"
export TERMINAIL_API_KEY="$TERMINAILI_API_KEY"
export OPENAI_MODEL="ipop/coder-balanced"

Core API

Choose the canonical model first, then add streaming or tool calls.

Choose a model

Start with ipop/coder-balanced for coding-agent and app quickstarts.

Model	Name	Context	Price	Recommended for	Lifecycle
`ipop/coder-balanced`	Terminaili coder balanced	1,000,000	catalog rate	Agentic coding	Recommended

Compatibility

Model	Status	Chat	Tools	JSON	Vision	Reasoning	Code	Streaming	Embeddings
`ipop/coder-balanced`	Ready	Yes	Yes	Yes	No	No	Yes	Yes	No

Recommendations and estimates

Model recommendations

curl https://api.terminaili.com/v1/model-recommendations \
  -H "Content-Type: application/json" \
  -d '{
    "task": "json schema extraction",
    "require_json_schema": true,
    "prefer_low_latency": true
  }'

Cost estimate

curl https://api.terminaili.com/v1/estimate \
  -H "Authorization: Bearer $TERMINAILI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "ipop/coder-balanced",
    "messages": [{ "role": "user", "content": "Estimate this request before running it." }],
    "max_tokens": 256
  }'

Streaming

Set stream: true to receive Server-Sent Events from chat completions.

Streaming request

{
  "model": "ipop/coder-balanced",
  "stream": true,
  "stream_options": { "include_usage": true },
  "messages": [{ "role": "user", "content": "Stream a short answer." }]
}

Tool Calls

Send OpenAI-compatible function definitions in tools.

Tool call request

{
  "model": "ipop/coder-balanced",
  "messages": [{ "role": "user", "content": "What is the weather in SF?" }],
  "tools": [
    {
      "type": "function",
      "function": {
        "name": "get_weather",
        "description": "Get current weather by city.",
        "parameters": {
          "type": "object",
          "properties": { "city": { "type": "string" } },
          "required": ["city"]
        }
      }
    }
  ]
}

Next steps

Open the deeper page only when you need reference tables, guides, or client setup.

Next steps

Use the split docs pages for details beyond the first working call.

API reference Guides Integrations OpenAPI JSON Plain text llms.txt