Quick start
Make the first traced API call.
Create a dashboard key, keep it out of client code, call the endpoint that matches the job, then verify the request in logs.
3. Send requestStart with the smallest working body.
4. Check logsConfirm status, usage, and request ID.Before you begin
Keep the key out of client code
Create a VIABLE Lab API key, then load it from your app's secret store, environment manager, or backend config. The examples use <VIABLE_API_KEY> as a placeholder.
Agent handoff
Using an Agent?
Paste this into Codex, Claude Code, Gemini CLI, Cursor, Windsurf, or a vibe-coding workflow so it starts from the maintained API guide.
CodexClaude CodeGemini CLICursorWindsurf
Open Agent guide Use the VIABLE Lab API guide at https://developers.viablelab.org/agents/viablelab-api.md. The API base URL is https://api.viablelab.org. Use a VIABLE Lab API key for protected API calls. Prefer documented endpoint shapes, full model IDs, and supported aliases.General
/v1/chat/completions
Educational
/v1/chat/vetting
Structured
/v1/vetting/full
Retrieval
/v1/embeddings
Media
/v1/images/generations
Audio
/v1/audio/*
First request
OpenAI Python SDK
OpenAI-compatible endpoints can use OpenAI SDKs by setting the base URL to https://api.viablelab.org/v1.
from openai import OpenAI
client = OpenAI(
base_url="https://api.viablelab.org/v1",
api_key="<VIABLE_API_KEY>",
)
response = client.chat.completions.create(
model="viable-2",
messages=[
{"role": "user", "content": "Explain how AI works in one sentence."}
],
)
print(response.choices[0].message.content)