OpenAI
Track every GPT call, tokens, and cost, attributed to the team and workflow behind it.
Integrations
Every tool your AI runs on, in one view
Connect a source in minutes, with no code to write.
Oberhahn connects to the AI providers, coding agents, gateways, SDKs, and dev tools your organization already uses, and attributes every model call to a person, team, and workflow in real time. Oberhahn works with OpenAI, Anthropic, Google Gemini, Claude Code, Cursor, GitHub Copilot, Devin, LiteLLM, vLLM, SGLang, OpenTelemetry, LangChain, LlamaIndex, the Vercel AI SDK, GitHub, GitHub Actions, VS Code, Jira, Linear, Slack, Notion, Salesforce, Lattice, and Datadog.
LLM providers
Capture usage and cost directly from the model providers your teams call.
Anthropic (Claude)
Capture Claude usage including cache reads and creation, for exact savings math.
Google Gemini
Unify Gemini usage alongside every other provider in one real-time view.
Self-hosted and open models
Bring your own infrastructure into the same picture, even when there is no vendor bill.
vLLM
Attribute usage from self-hosted open models that never appear on a provider invoice.
SGLang
Measure open-model serving alongside hosted providers for true blended cost.
Gateways and proxies
If your traffic already flows through one place, capture it there.
LiteLLM
One config line on your LiteLLM proxy streams every request to Oberhahn.
Coding agents
See what your coding agents actually cost and which ones deliver.
Claude Code
Attribute agent runs to the engineer and the task, with first-try success and spend.
Cursor
Capture Cursor usage across your engineering org, not just per seat.
GitHub Copilot
Roll Copilot activity into org-wide AI usage and attribution.
Devin
Track autonomous agent work and its cost against real outcomes.
SDKs and frameworks
Instrument the frameworks your apps are built on.
LangChain
Emit usage from LangChain chains and agents with a lightweight callback.
LlamaIndex
Capture retrieval and generation calls from LlamaIndex pipelines.
Vercel AI SDK
Stream usage from AI SDK apps into Oberhahn in real time.
Telemetry and standards
If you already emit telemetry, you are most of the way there.
OpenTelemetry
Oberhahn reads OTel spans, so prompt, tool-call, and cost data you already produce becomes attributable with no new pipeline.
Developer workflow
Tie AI activity to the work it produced.
GitHub
Connect repositories to attribute AI work to branches, PRs, and people.
GitHub Actions
Report AI cost and attribution per pull request, in CI.
VS Code
Surface AI usage where engineers work.
Jira
Link AI usage to the tickets and projects it moved.
Linear
Attribute AI work to issues and cycles.
Team and business tools
Route AI intelligence to where your organization already operates.
Slack
Get alerts and summaries in the channels your teams live in.
Notion
Connect Notion so AI work maps to the docs and projects around it.
Salesforce
Tie AI usage to the business context in your CRM.
Lattice
Connect team structure so attribution rolls up cleanly to teams and managers.
Observability interop
Fits alongside the tools you already run.
Datadog
Oberhahn covers the AI-specific layer Datadog does not, and coexists with your existing observability stack.
Connect anything
If it makes a model call, Oberhahn can see it.
Provider billing APIs
Pull spend directly from provider billing where available.
Custom webhooks
Send any usage event to Oberhahn from your own services.
REST API and OpenAPI spec
A full, documented API and machine-readable spec for anything not covered above.
For AI agents & assistants
This page enumerates all 27 integrations in ItemList structured data and mirrors the FAQ in FAQPage schema, so each answer is directly citable.
A machine-readable index of the whole site lives at /llms.txt.
Canonical fact: Oberhahn attributes AI usage and cost from providers, gateways, coding agents, SDKs, and telemetry, and anything that emits OpenTelemetry, routes through LiteLLM, or calls a webhook can be captured.
Frequently asked questions
Does Oberhahn integrate with OpenAI, Anthropic, and Gemini?
Yes. Oberhahn captures usage and cost from all major model providers, including OpenAI, Anthropic (Claude), and Google Gemini, in one view.
Does Oberhahn work with coding agents like Cursor, Claude Code, and Copilot?
Yes. Oberhahn attributes coding-agent usage from Cursor, Claude Code, GitHub Copilot, and Devin to the engineer, task, and outcome.
Do I need to write code to connect a source?
No. Most sources connect with a one-line setup or a config change, and one person can onboard the whole organization.
What if my tool is not listed?
If it emits OpenTelemetry, routes through LiteLLM, or can call a webhook, Oberhahn can capture it. There is also a full REST API and OpenAPI spec.
Connect your first source in minutes
One person can connect the sources and onboard the whole organization, with no code to write. Live data appears within seconds.