Responses API

Access compatible provider Responses endpoints without losing provider-specific fields.

Use responses() for providers that expose an OpenAI-compatible Responses API.

import { responses } from "any-llm-ts";

const response = await responses({
  provider: "openai",
  model: "gpt-4.1-mini",
  input: "Explain event loops in one paragraph.",
});

The reusable client exposes the same operation:

const response = await llm.responses({
  model: "gpt-4.1-mini",
  input: [
    {
      role: "user",
      content: [{ type: "input_text", text: "Hello" }],
    },
  ],
});

ResponsesParams intentionally allows additional keys, and the result is currently unknown. The Responses API evolves quickly and has a much larger provider-specific surface than chat completions; this design avoids throwing away fields or publishing an inaccurate lowest-common- denominator type.

Narrow the result in application code or use the relevant provider SDK's response type:

const response = await llm.responses(params);

if (typeof response === "object" && response !== null && "output" in response) {
  // Work with the provider response after validation.
}

Streaming

const events = await llm.responses({
  model: "gpt-4.1-mini",
  input: "Hello",
  stream: true,
});

for await (const event of events) {
  // Validate and handle provider response events.
}

Check metadata.capabilities.responses first. Unsupported providers reject with UnsupportedOperationError.

Streaming

On this page