Skip to content

RunForge Documentation

Monitor and optimize your AI usage. Private by default. Built-in dashboards, alerts, and experiments to keep costs down and quality up.

title: RunForge Documentation status: current lastReviewed: 2025-08-17

Start here

Track LLM calls in 2 lines

Easily instrument your app without changing your code paths.

import OpenAI from 'openai'
import { withRunForge } from '../sdk-ts/index'

const openai = withRunForge(new OpenAI({ apiKey: process.env.OPENAI_API_KEY }))
const res = await openai.chat.completions.create({
  model: 'gpt-4o-mini',
  messages: [{ role: 'user', content: 'Hello!' }],
})

from openai import OpenAI
from runforge import with_runforge
import os

client = with_runforge(OpenAI(api_key=os.environ['OPENAI_API_KEY']))
res = client.chat.completions.create(model='gpt-4o-mini', messages=[{"role":"user","content":"Hello!"}])

The wrapper auto‑extracts tokens, latency, and cost, and sends only usage metadata to /api/ingest (never prompts/outputs).

Who is this for?

  • đź‘Ą End users: Track costs, latency, and reliability in the User Guide
  • 👨‍💻 Developers: Set up in minutes with the Quickstart and SDKs (TypeScript, Python)

Highlights

  • Real‑time runs via Convex telemetry
  • Durable analytics in PostgreSQL (Prisma)
  • Experiments, alerts, and usage dashboards
  • Privacy by design: no prompts or outputs ingested

See also: Overview, Data model, API reference

License

  • Core: Apache-2.0 (see /LICENSE)
  • SDKs: MIT (see /sdk-ts/LICENSE and /sdk-py/LICENSE)