AI gateway

AI gateway carries every model call a Startup makes. It picks the model by Policy, checks cost Limits before the call, and keeps the full request and response.

ContractCopy Anchor Link

• Four operations: generate, embed, classify and transcribe. Every Agent, Function and Evaluator calls models through them.
• Code names a task class, never a model: classify, judge, draft, agent, embed or transcribe.
• A Policy maps each class to a ranked list of models, cheapest capable first. Policies inherit from Studio to Startup to call, like a Setting.
• Changing a model is a Change to a Policy, never to code.
• A call over its cost Limit waits for an Approval.
• Every request and response is kept forever, with its tokens, latency, Cost and the Task it served.
• Results stream as they generate.
• A Policy can split by Flag, so a cheaper model runs as an Experiment, judged on cost per accepted outcome.
• AI models explains how a Startup chooses and grades its models.

VerbsCopy Anchor Link

None.

InterfacesCopy Anchor Link

• API
• SDK
• CLI
• MCP

ExampleCopy Anchor Link

// replydesk.io answers support email for small online shops. Its agent names task classes, never models.
import { $ } from 'startup-builder'

await $.Agent.create({
  instructions: 'Label each email refund, shipping or other, then draft the reply.',
  model: 'draft',
  tools: ['Conversation.answer'],
})