IRChatResponse

Defined in: packages/ai.matey.types/src/ir.ts:795

Universal chat completion response.

Normalized format that all backend adapters convert to and all frontend adapters consume.

Example

const response: IRChatResponse = {
  message: {
    role: 'assistant',
    content: 'Hello! How can I help you today?'
  },
  finishReason: 'stop',
  usage: {
    promptTokens: 15,
    completionTokens: 10,
    totalTokens: 25
  },
  metadata: {
    requestId: 'req_abc123',
    timestamp: Date.now(),
    provenance: {
      frontend: 'openai',
      backend: 'openai'
    }
  }
};

Properties

finishReason

readonly finishReason: FinishReason

Defined in: packages/ai.matey.types/src/ir.ts:804

Why generation finished.

---

message

readonly message: IRMessage

Defined in: packages/ai.matey.types/src/ir.ts:799

Generated message from the assistant.

---

metadata

readonly metadata: IRMetadata

Defined in: packages/ai.matey.types/src/ir.ts:815

Response metadata. Includes original request metadata plus backend provenance.

---

raw?

readonly optional raw: Record<string, unknown>

Defined in: packages/ai.matey.types/src/ir.ts:820

Provider-specific response data.

---

usage?

readonly optional usage: IRUsage

Defined in: packages/ai.matey.types/src/ir.ts:809

Token usage statistics.