ExducerAI
MCP server by ExducerAI · mcp
ExducerAI rents whole GPUs in Montréal (CA-QC) by the hour and serves an OpenAI-compatible Models API. Amounts are in CAD. Read prices, servers and templates from the tools; never from memory. A booking is a request: it runs later (wait_for_deployment). Billing is per minute while it runs. Never call create_booking, extend_deployment, set_auto_renew or stop_deployment without the user's explicit yes; quote first. A stop deletes the deployment's disk. A key that starts with tml_test_ is a sandbox key: everything is simulated and nothing is charged. To run commands on a deployment use SSH (the CLI: turbo ssh ID --batch -- COMMAND). Docs: search_docs.
Endpoint: https://exducer.ai/mcp
Verification: observed. Source: crawl.
Capabilities (22)
- get_pricing
- check_availability
- get_service_status
- list_servers
- list_templates
- quote_booking
- search_docs
- read_doc
- get_account
- list_deployments
- get_deployment
- get_deployment_events
- wait_for_deployment
- get_billing
- get_usage
- list_invoices
- list_ssh_keys
- get_storage_usage
- create_booking
- extend_deployment
- set_auto_renew
- stop_deployment