Martin · Model router
Available nowOne API for model access, credential management and credit controls
Martin routes your model calls through a single OpenAI-compatible API, manages credentials in one Token Studio and caps spend by API key, agent and team
key/batch-eval is running at 4.2× its seven-day average
01 — Model access
Point your client at Martin. Keep your code
Martin is OpenAI-compatible. Change the base URL and API key, and you can call models from multiple providers through one endpoint
We do not publish the provider list yet. Book a demo to see which models fit your use case
client = OpenAI( api_key=os.environ["OPENAI_API_KEY"], base_url=os.environ["GOSEVEN_BASE_URL"], api_key=os.environ["GOSEVEN_API_KEY"],)02 — Credentials and credit controls
Limits that enforce themselves
When a key or team exceeds its limit, Martin acts. Nobody has to watch a dashboard
-
01
Issue keys
Create an API key for each team and application from one GoSeven account
-
02
Set limits
Assign a credit limit to each API key and team
-
03
Freeze within 24 hours
Martin freezes an over-limit key or team within 24 hours and holds its credits
-
04
Alert the owner
The account owner receives an alert and decides what happens next
03 — Eliminate waste
Stop paying for calls you don’t need
Call optimisation
Martin optimises how calls are made, so the same work costs less
Anomaly detection
Martin flags unusual consumption early, by API key and by team
04 — Data and billing
Your prompts stay yours
- Prompts and responses
- Not stored, or privacy-processed
- Billing
- Prepaid credits for the routing service