pruneMessages()

The pruneMessages function is used to prune or filter an array of ModelMessage objects. This is useful for reducing message context (to save tokens), removing intermediate reasoning, or trimming tool calls and empty messages before sending to an LLM.

1
import { pruneMessages, streamText } from 'ai';
2

3
export async function POST(req: Request) {
4
  const { messages } = await req.json();
5

6
  const prunedMessages = pruneMessages({
7
    messages,
8
    reasoning: 'before-last-message',
9
    toolCalls: 'before-last-2-messages',
10
    emptyMessages: 'remove',
11
  });
12

13
  const result = streamText({
14
    model: "anthropic/claude-sonnet-4.5",
15
    messages: prunedMessages,
16
  });
17

18
  return result.toUIMessageStreamResponse();
19
}

Import

import { pruneMessages } from "ai"

API Signature

Parameters

Returns

An array of ModelMessage objects, pruned according to the provided options.

Example Usage

1
import { pruneMessages } from 'ai';
2

3
const pruned = pruneMessages({
4
  messages,
5
  reasoning: 'all', // Remove all reasoning parts
6
  toolCalls: 'before-last-message', // Remove tool calls except those in the last message
7
});

Pruning Options

• reasoning: Removes reasoning parts from assistant messages. Use 'all' to remove all, 'before-last-message' to keep reasoning in the last message, or 'none' to retain all reasoning.
• toolCalls: Prune tool-call, tool-result, and tool-approval chunks from assistant/tool messages. Options include:
  • 'all': Prune all such content.
  • 'before-last-message': Prune except in the last message.
  • before-last-N-messages: Prune except in the last N messages.
  • 'none': Do not prune.
  • Or provide an array for per-tool fine control.
• emptyMessages: Set to 'remove' (default) to exclude messages that have no content after pruning.

Tip: pruneMessages is typically used prior to sending a context window to an LLM to reduce message/token count, especially after a series of tool-calls and approvals.

For advanced usage and the full list of possible message parts, see ModelMessage and pruneMessages implementation.