Examples

Any client that supports a custom OpenAI base URL works. Set:

• Base URL → https://<host>/api/openai/v1 (no trailing path beyond /v1)
• API key → your sk-tbl-… key
• Model → prefer default-assistant or an agent slug from GET /models

OpenAI Python SDK

from openai import OpenAI

client = OpenAI(
    api_key="sk-tbl-...",
    base_url="https://taibles.ai/api/openai/v1",
)

# Agent (recommended starting point)
r = client.chat.completions.create(
    model="default-assistant",
    messages=[{"role": "user", "content": "Draft a status update for the ops channel."}],
)
print(r.choices[0].message.content)

# Stream
with client.chat.completions.create(
    model="default-assistant",
    messages=[{"role": "user", "content": "Explain our refund policy briefly."}],
    stream=True,
) as stream:
    for chunk in stream:
        delta = chunk.choices[0].delta.content or ""
        print(delta, end="", flush=True)

OpenAI Node.js SDK

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.TAIBLES_API_KEY,
  baseURL: "https://taibles.ai/api/openai/v1",
});

const stream = await client.chat.completions.create({
  model: "default-assistant",
  messages: [{ role: "user", content: "List three next actions for onboarding." }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || "");
}

Continue a conversation

first = client.chat.completions.create(
    model="default-assistant",
    messages=[{"role": "user", "content": "My project codename is Orion."}],
)
conversation_id = first.taibles["conversation_id"]  # or response header

follow_up = client.chat.completions.create(
    model="default-assistant",
    messages=[{"role": "user", "content": "What was the codename?"}],
    extra_body={"taibles": {"conversation_id": conversation_id}},
)

Depending on the SDK version, extension fields may appear as a dict on the response object or only in the raw JSON / X-Taibles-Conversation-Id header. Prefer the header if your client strips unknown fields.

List models

curl "$BASE_URL/models" \
  -H "Authorization: Bearer $YOUR_API_KEY"

Use returned agent id values as model for specialized agents you built in TAIBLES.

Cursor / other OpenAI-compatible tools

In any tool that asks for an OpenAI-compatible provider:

Setting	Value
API Base / Base URL	https://taibles.ai/api/openai/v1
API Key	sk-tbl-…
Model	default-assistant (or an agent slug / LLM URL from /models)

You are calling your TAIBLES workspace — not a generic public model marketplace. Capabilities follow the agent you select.