MVP: docs on withohm.dev · chat edge = local :8081 (or your deploy) · api.withohm.dev returns edge-pending until AWS cutover. Status

Drop-in examples

Ohm’s distribution channel is the OpenAI-compatible API. No new protocol.

Keys use legacy prefix sk-at-…. Prefer local edge until public API cutover.

Local (supported MVP)

from openai import OpenAI
client = OpenAI(base_url="http://localhost:8081/v1", api_key="sk-at-dev")
import OpenAI from "openai";
const client = new OpenAI({
  apiKey: "sk-at-dev",
  baseURL: "http://127.0.0.1:8081/v1",
});
curl -s http://localhost:8081/v1/chat/completions \
  -H "Authorization: Bearer sk-at-dev" \
  -H "Content-Type: application/json" \
  -d '{"model":"mock","messages":[{"role":"user","content":"hi"}]}'

Production host (reserved)

https://api.withohm.dev/v1 is the documented public hostname (ACM issued). Chat traffic is not a public SLA until the AWS edge is cut over — do not treat HTML on that host as the API.

Web context (compliance required)

Requires web_purpose and web_compliance_ack, plus Terms/DPA acks. See legal.

curl -s http://localhost:8081/v1/chat/completions \
  -H "Authorization: Bearer sk-at-dev" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Summarize"}],"fetch_web_context":true,"web_urls":["https://example.com"],"web_purpose":"business_catalog","web_compliance_ack":true,"terms_ack":true,"dpa_ack":true}'