streams/openai-stream: allow returning strings in function callback (#…

…326)
vercel · Jul 13, 2023 · e361114 · e361114
1 parent 8ca6e39
commit e361114
Show file tree

Hide file tree

Showing 4 changed files with 12 additions and 3 deletions.
diff --git a/.changeset/sharp-hairs-itch.md b/.changeset/sharp-hairs-itch.md
@@ -0,0 +1,5 @@
+---
+'ai': patch
+---
+
+OpenAI functions: allow returning string in callback
diff --git a/docs/pages/docs/guides/functions.mdx b/docs/pages/docs/guides/functions.mdx
@@ -66,6 +66,7 @@ You can then choose how you want to handle each function call: on the server or
 
 On the server, you can pass an `experimental_onFunctionCall` callback to the `OpenAIStream`, which will be called when the model calls a function.
 In order to support recursively calling functions and to construct the message context in the nested OpenAI calls, you can use `createFunctionCallMessages` to get the "assistant" and "function" messages.
+You can also return a string which will be sent to the client as the "assistant" message (or returned back to the model as a response to a recursive function call).
 
 ```ts {2,3,4,15,17,21,22}
 const stream = OpenAIStream(response, {

diff --git a/packages/core/streams/ai-stream.ts b/packages/core/streams/ai-stream.ts
@@ -4,7 +4,6 @@ import {
   type ParsedEvent,
   type ReconnectInterval
 } from 'eventsource-parser'
-import { CreateMessage } from '../shared/types'
 
 export interface FunctionCallPayload {
   name: string

diff --git a/packages/core/streams/openai-stream.ts b/packages/core/streams/openai-stream.ts
@@ -31,7 +31,7 @@ export type OpenAIStreamCallbacks = AIStreamCallbacks & {
    *     // ... run your custom logic here
    *     const result = await myFunction(functionCallPayload)
    *
-   *     // Ask for another completion
+   *     // Ask for another completion, or return a string to send to the client as an assistant message.
    *     return await openai.createChatCompletion({
    *       model: 'gpt-3.5-turbo-0613',
    *       stream: true,
@@ -48,7 +48,7 @@ export type OpenAIStreamCallbacks = AIStreamCallbacks & {
     createFunctionCallMessages: (
       functionCallResult: JSONValue
     ) => CreateMessage[]
-  ) => Promise<Response | undefined>
+  ) => Promise<Response | undefined | void | string>
 }
 
 /**
@@ -270,6 +270,10 @@ function createFunctionCallTransformer(
           // so we just return the function call as a message
           controller.enqueue(textEncoder.encode(aggregatedResponse))
           return
+        } else if (typeof functionResponse === 'string') {
+          // The user returned a string, so we just return it as a message
+          controller.enqueue(textEncoder.encode(functionResponse))
+          return
         }
 
         // Recursively