Fine-grained tool streaming

Stream tool inputs without server-side JSON buffering for latency-sensitive applications.

Note

This feature is eligible for Zero Data Retention (ZDR). When your organization has a ZDR arrangement, data sent through this feature is not stored after the API response is returned.

Fine-grained tool streaming is available on all models and all platforms. It enables streaming of tool use parameter values without buffering or JSON validation, reducing the latency to begin receiving large parameters.

Warning

When using fine-grained tool streaming, you may potentially receive invalid or partial JSON inputs. Make sure to account for these edge cases in your code.

How to use fine-grained tool streaming

Fine-grained tool streaming is supported on the Claude API, Claude Platform on AWS, Amazon Bedrock, Vertex AI, and Microsoft Foundry. To use it, set eager_input_streaming to true on any user-defined tool where you want fine-grained streaming enabled, and enable streaming on your request.

Here's an example of how to use fine-grained tool streaming with the API:

curl https://api.anthropic.com/v1/messages \
  -H "content-type: application/json" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-opus-4-7",
    "max_tokens": 65536,
    "tools": [
      {
        "name": "make_file",
        "description": "Write text to a file",
        "eager_input_streaming": true,
        "input_schema": {
          "type": "object",
          "properties": {
            "filename": {
              "type": "string",
              "description": "The filename to write text to"
            },
            "lines_of_text": {
              "type": "array",
              "description": "An array of lines of text to write to the file"
            }
          },
          "required": ["filename", "lines_of_text"]
        }
      }
    ],
    "messages": [
      {
        "role": "user",
        "content": "Can you write a long poem and make a file called poem.txt?"
      }
    ],
    "stream": true
  }'

ant messages create --stream --format jsonl <<'YAML' |
model: claude-opus-4-7
max_tokens: 65536
tools:
  - name: make_file
    description: Write text to a file
    eager_input_streaming: true
    input_schema:
      type: object
      properties:
        filename:
          type: string
          description: The filename to write text to
        lines_of_text:
          type: array
          description: An array of lines of text to write to the file
      required:
        - filename
        - lines_of_text
messages:
  - role: user
    content: Can you write a long poem and make a file called poem.txt?
YAML
  jq 'select(.type == "message_delta") | .usage'

import anthropic

client = anthropic.Anthropic()

with client.messages.stream(
    max_tokens=65536,
    model="claude-opus-4-7",
    tools=[
        {
            "name": "make_file",
            "description": "Write text to a file",
            "eager_input_streaming": True,
            "input_schema": {
                "type": "object",
                "properties": {
                    "filename": {
                        "type": "string",
                        "description": "The filename to write text to",
                    },
                    "lines_of_text": {
                        "type": "array",
                        "description": "An array of lines of text to write to the file",
                    },
                },
                "required": ["filename", "lines_of_text"],
            },
        }
    ],
    messages=[
        {
            "role": "user",
            "content": "Can you write a long poem and make a file called poem.txt?",
        }
    ],
) as stream:
    final_message = stream.get_final_message()

print(f"Input tokens: {final_message.usage.input_tokens}")
print(f"Output tokens: {final_message.usage.output_tokens}")

import Anthropic from "@anthropic-ai/sdk";

const anthropic = new Anthropic();

const stream = anthropic.messages.stream({
  model: "claude-opus-4-7",
  max_tokens: 65536,
  tools: [
    {
      name: "make_file",
      description: "Write text to a file",
      eager_input_streaming: true,
      input_schema: {
        type: "object",
        properties: {
          filename: {
            type: "string",
            description: "The filename to write text to"
          },
          lines_of_text: {
            type: "array",
            description: "An array of lines of text to write to the file"
          }
        },
        required: ["filename", "lines_of_text"]
      }
    }
  ],
  messages: [
    {
      role: "user",
      content: "Can you write a long poem and make a file called poem.txt?"
    }
  ]
});

const message = await stream.finalMessage();
console.log(`Input tokens: ${message.usage.input_tokens}`);
console.log(`Output tokens: ${message.usage.output_tokens}`);

using System.Text.Json;
using Anthropic;
using Anthropic.Models.Messages;

AnthropicClient client = new();

MessageCreateParams parameters = new()
{
    Model = Model.ClaudeOpus4_7,
    MaxTokens = 65536,
    Tools =
    [
        new Tool
        {
            Name = "make_file",
            Description = "Write text to a file",
            EagerInputStreaming = true,
            InputSchema = new InputSchema
            {
                Properties = new Dictionary<string, JsonElement>
                {
                    ["filename"] = JsonSerializer.SerializeToElement(
                        new { type = "string", description = "The filename to write text to" }
                    ),
                    ["lines_of_text"] = JsonSerializer.SerializeToElement(
                        new { type = "array", description = "An array of lines of text to write to the file" }
                    ),
                },
                Required = ["filename", "lines_of_text"],
            },
        },
    ],
    Messages =
    [
        new()
        {
            Role = Role.User,
            Content = "Can you write a long poem and make a file called poem.txt?",
        },
    ],
};

long inputTokens = 0;
long outputTokens = 0;

await foreach (var streamEvent in client.Messages.CreateStreaming(parameters))
{
    switch (streamEvent.Value)
    {
        case RawMessageStartEvent startEvent:
            inputTokens = startEvent.Message.Usage.InputTokens;
            break;
        case RawMessageDeltaEvent deltaEvent:
            outputTokens = deltaEvent.Usage.OutputTokens;
            break;
    }
}

Console.WriteLine({{CONTENT}}quot;Input tokens: {inputTokens}");
Console.WriteLine({{CONTENT}}quot;Output tokens: {outputTokens}");

package main

import (
	"context"
	"fmt"

	"github.com/anthropics/anthropic-sdk-go"
)

func main() {
	client := anthropic.NewClient()

	makeFileTool := anthropic.ToolParam{
		Name:                "make_file",
		Description:         anthropic.String("Write text to a file"),
		EagerInputStreaming: anthropic.Bool(true),
		InputSchema: anthropic.ToolInputSchemaParam{
			Properties: map[string]any{
				"filename": map[string]any{
					"type":        "string",
					"description": "The filename to write text to",
				},
				"lines_of_text": map[string]any{
					"type":        "array",
					"description": "An array of lines of text to write to the file",
				},
			},
			Required: []string{"filename", "lines_of_text"},
		},
	}

	stream := client.Messages.NewStreaming(context.Background(), anthropic.MessageNewParams{
		Model:     anthropic.ModelClaudeOpus4_7,
		MaxTokens: 65536,
		Tools:     []anthropic.ToolUnionParam{{OfTool: &makeFileTool}},
		Messages: []anthropic.MessageParam{
			anthropic.NewUserMessage(anthropic.NewTextBlock(
				"Can you write a long poem and make a file called poem.txt?",
			)),
		},
	})

	message := anthropic.Message{}
	for stream.Next() {
		event := stream.Current()
		if err := message.Accumulate(event); err != nil {
			panic(err)
		}
	}
	if err := stream.Err(); err != nil {
		panic(err)
	}

	fmt.Printf("Input tokens: %d\n", message.Usage.InputTokens)
	fmt.Printf("Output tokens: %d\n", message.Usage.OutputTokens)
}

import com.anthropic.client.AnthropicClient;
import com.anthropic.client.okhttp.AnthropicOkHttpClient;
import com.anthropic.core.JsonValue;
import com.anthropic.core.http.StreamResponse;
import com.anthropic.helpers.MessageAccumulator;
import com.anthropic.models.messages.MessageCreateParams;
import com.anthropic.models.messages.Model;
import com.anthropic.models.messages.RawMessageStreamEvent;
import com.anthropic.models.messages.Tool;
import com.anthropic.models.messages.Usage;

void main() {
    AnthropicClient client = AnthropicOkHttpClient.fromEnv();

    Tool makeFileTool = Tool.builder()
        .name("make_file")
        .description("Write text to a file")
        .eagerInputStreaming(true)
        .inputSchema(Tool.InputSchema.builder()
            .properties(Tool.InputSchema.Properties.builder()
                .putAdditionalProperty("filename", JsonValue.from(Map.of(
                    "type", "string",
                    "description", "The filename to write text to")))
                .putAdditionalProperty("lines_of_text", JsonValue.from(Map.of(
                    "type", "array",
                    "description", "An array of lines of text to write to the file")))
                .build())
            .addRequired("filename")
            .addRequired("lines_of_text")
            .build())
        .build();

    MessageCreateParams params = MessageCreateParams.builder()
        .model(Model.CLAUDE_OPUS_4_7)
        .maxTokens(65536L)
        .addTool(makeFileTool)
        .addUserMessage("Can you write a long poem and make a file called poem.txt?")
        .build();

    MessageAccumulator accumulator = MessageAccumulator.create();

    try (StreamResponse<RawMessageStreamEvent> streamResponse =
            client.messages().createStreaming(params)) {
        streamResponse.stream().forEach(accumulator::accumulate);
    }

    Usage usage = accumulator.message().usage();
    IO.println("Input tokens: " + usage.inputTokens());
    IO.println("Output tokens: " + usage.outputTokens());
}

<?php

use Anthropic\Client;
use Anthropic\Messages\Model;
use Anthropic\Messages\RawMessageDeltaEvent;
use Anthropic\Messages\RawMessageStartEvent;

$client = new Client();

$stream = $client->messages->createStream(
    maxTokens: 65536,
    model: Model::CLAUDE_OPUS_4_7,
    tools: [
        [
            'name' => 'make_file',
            'description' => 'Write text to a file',
            'eager_input_streaming' => true,
            'input_schema' => [
                'type' => 'object',
                'properties' => [
                    'filename' => [
                        'type' => 'string',
                        'description' => 'The filename to write text to',
                    ],
                    'lines_of_text' => [
                        'type' => 'array',
                        'description' => 'An array of lines of text to write to the file',
                    ],
                ],
                'required' => ['filename', 'lines_of_text'],
            ],
        ],
    ],
    messages: [
        [
            'role' => 'user',
            'content' => 'Can you write a long poem and make a file called poem.txt?',
        ],
    ],
);

$inputTokens = 0;
$outputTokens = 0;

foreach ($stream as $event) {
    if ($event instanceof RawMessageStartEvent) {
        $inputTokens = $event->message->usage->inputTokens;
    } elseif ($event instanceof RawMessageDeltaEvent) {
        $outputTokens = $event->usage->outputTokens;
    }
}

echo "Input tokens: {$inputTokens}\n";
echo "Output tokens: {$outputTokens}\n";

require "anthropic"

anthropic = Anthropic::Client.new

stream = anthropic.messages.stream(
  model: Anthropic::Models::Model::CLAUDE_OPUS_4_7,
  max_tokens: 65_536,
  tools: [
    {
      name: "make_file",
      description: "Write text to a file",
      eager_input_streaming: true,
      input_schema: {
        type: "object",
        properties: {
          filename: {
            type: "string",
            description: "The filename to write text to"
          },
          lines_of_text: {
            type: "array",
            description: "An array of lines of text to write to the file"
          }
        },
        required: ["filename", "lines_of_text"]
      }
    }
  ],
  messages: [
    {
      role: "user",
      content: "Can you write a long poem and make a file called poem.txt?"
    }
  ]
)

usage = stream.accumulated_message.usage
puts "Input tokens: #{usage.input_tokens}"
puts "Output tokens: #{usage.output_tokens}"

In this example, fine-grained tool streaming enables Claude to stream the lines of a long poem into the tool call make_file without buffering to validate if the lines_of_text parameter is valid JSON. This means you can see the parameter stream as it arrives, without having to wait for the entire parameter to buffer and validate.

Note

With fine-grained tool streaming, tool input chunks start arriving sooner because the server skips JSON-validation buffering. Chunks are typically longer and contain fewer mid-token breaks as a side effect.

Warning

Because fine-grained streaming sends parameters without buffering or JSON validation, there is no guarantee that the resulting stream will complete in a valid JSON string. Particularly, if the stop reason max_tokens is reached, the stream may end midway through a parameter and may be incomplete. You generally have to write specific support to handle when max_tokens is reached.

Accumulating tool input deltas

When a tool_use content block streams, the initial content_block_start event contains input: {} (an empty object). This is a placeholder. The actual input arrives as a series of input_json_delta events, each carrying a partial_json string fragment. To assemble the full input, concatenate these fragments and parse the result when the block closes.

Where your SDK provides an accumulator helper (as used in the first example on this page), it handles this for you. The manual pattern is for SDKs without a helper, or when you need to react to partial input before the block closes.

The accumulation contract:

On content_block_start with type: "tool_use", initialize an empty string: input_json = ""
For each content_block_delta with type: "input_json_delta", append: input_json += event.delta.partial_json
On content_block_stop, parse the accumulated string: json.loads(input_json)

The type mismatch between the initial input: {} (object) and partial_json (string) is by design. The empty object marks the slot in the content array; the delta strings build the real value.

import json
import anthropic

client = anthropic.Anthropic()

tool_inputs: dict[int, str] = {}  # index -> accumulated JSON string

with client.messages.stream(
    model="claude-opus-4-7",
    max_tokens=1024,
    tools=[
        {
            "name": "get_weather",
            "description": "Get current weather for a city",
            "eager_input_streaming": True,
            "input_schema": {
                "type": "object",
                "properties": {"city": {"type": "string"}},
                "required": ["city"],
            },
        }
    ],
    messages=[{"role": "user", "content": "Weather in Paris?"}],
) as stream:
    for event in stream:
        match event.type:
            case "content_block_start" if event.content_block.type == "tool_use":
                tool_inputs[event.index] = ""
            case "content_block_delta" if event.delta.type == "input_json_delta":
                tool_inputs[event.index] += event.delta.partial_json
            case "content_block_stop" if event.index in tool_inputs:
                parsed = json.loads(tool_inputs[event.index])
                print(f"Tool input: {parsed}")

import Anthropic from "@anthropic-ai/sdk";

const anthropic = new Anthropic();

const toolInputs = new Map<number, string>();

const stream = anthropic.messages.stream({
  model: "claude-opus-4-7",
  max_tokens: 1024,
  tools: [
    {
      name: "get_weather",
      description: "Get current weather for a city",
      eager_input_streaming: true,
      input_schema: {
        type: "object",
        properties: { city: { type: "string" } },
        required: ["city"]
      }
    }
  ],
  messages: [{ role: "user", content: "Weather in Paris?" }]
});

for await (const event of stream) {
  if (event.type === "content_block_start" && event.content_block.type === "tool_use") {
    toolInputs.set(event.index, "");
  } else if (event.type === "content_block_delta" && event.delta.type === "input_json_delta") {
    toolInputs.set(
      event.index,
      (toolInputs.get(event.index) ?? "") + event.delta.partial_json
    );
  } else if (event.type === "content_block_stop" && toolInputs.has(event.index)) {
    const parsed = JSON.parse(toolInputs.get(event.index)!);
    console.log("Tool input:", parsed);
  }
}

using System.Text;
using System.Text.Json;
using Anthropic;
using Anthropic.Models.Messages;

AnthropicClient client = new();

MessageCreateParams parameters = new()
{
    Model = Model.ClaudeOpus4_7,
    MaxTokens = 1024,
    Tools =
    [
        new Tool
        {
            Name = "get_weather",
            Description = "Get current weather for a city",
            EagerInputStreaming = true,
            InputSchema = new InputSchema
            {
                Properties = new Dictionary<string, JsonElement>
                {
                    ["city"] = JsonSerializer.SerializeToElement(new { type = "string" }),
                },
                Required = ["city"],
            },
        },
    ],
    Messages = [new() { Role = Role.User, Content = "Weather in Paris?" }],
};

// Block index -> accumulated JSON fragments
// The C# SDK does not currently provide a stream accumulator for tool input;
// the manual pattern shown here is the supported approach.
var toolInputs = new Dictionary<long, StringBuilder>();

await foreach (var streamEvent in client.Messages.CreateStreaming(parameters))
{
    if (
        streamEvent.TryPickContentBlockStart(out var start)
        && start.ContentBlock.TryPickToolUse(out _)
    )
    {
        toolInputs[start.Index] = new StringBuilder();
    }
    else if (
        streamEvent.TryPickContentBlockDelta(out var delta)
        && delta.Delta.TryPickInputJson(out var inputJson)
    )
    {
        toolInputs[delta.Index].Append(inputJson.PartialJson);
    }
    else if (
        streamEvent.TryPickContentBlockStop(out var stop)
        && toolInputs.TryGetValue(stop.Index, out var accumulated)
    )
    {
        using var parsed = JsonDocument.Parse(accumulated.ToString());
        Console.WriteLine({{CONTENT}}quot;Tool input: {parsed.RootElement}");
    }
}

package main

import (
	"context"
	"encoding/json"
	"fmt"

	"github.com/anthropics/anthropic-sdk-go"
)

func main() {
	client := anthropic.NewClient()

	toolInputs := map[int64]string{} // content block index -> accumulated JSON

	stream := client.Messages.NewStreaming(context.Background(), anthropic.MessageNewParams{
		Model:     anthropic.ModelClaudeOpus4_7,
		MaxTokens: 1024,
		Tools: []anthropic.ToolUnionParam{{
			OfTool: &anthropic.ToolParam{
				Name:                "get_weather",
				Description:         anthropic.String("Get current weather for a city"),
				EagerInputStreaming: anthropic.Bool(true),
				InputSchema: anthropic.ToolInputSchemaParam{
					Properties: map[string]any{
						"city": map[string]any{"type": "string"},
					},
					Required: []string{"city"},
				},
			},
		}},
		Messages: []anthropic.MessageParam{
			anthropic.NewUserMessage(anthropic.NewTextBlock("Weather in Paris?")),
		},
	})

	for stream.Next() {
		switch event := stream.Current().AsAny().(type) {
		case anthropic.ContentBlockStartEvent:
			if _, ok := event.ContentBlock.AsAny().(anthropic.ToolUseBlock); ok {
				toolInputs[event.Index] = ""
			}
		case anthropic.ContentBlockDeltaEvent:
			if delta, ok := event.Delta.AsAny().(anthropic.InputJSONDelta); ok {
				toolInputs[event.Index] += delta.PartialJSON
			}
		case anthropic.ContentBlockStopEvent:
			if accumulated, ok := toolInputs[event.Index]; ok {
				var parsed map[string]any
				if err := json.Unmarshal([]byte(accumulated), &parsed); err != nil {
					panic(err)
				}
				fmt.Println("Tool input:", parsed)
			}
		}
	}
	if err := stream.Err(); err != nil {
		panic(err)
	}
}

import com.anthropic.client.AnthropicClient;
import com.anthropic.client.okhttp.AnthropicOkHttpClient;
import com.anthropic.core.JsonValue;
import com.anthropic.core.http.StreamResponse;
import com.anthropic.models.messages.MessageCreateParams;
import com.anthropic.models.messages.Model;
import com.anthropic.models.messages.RawMessageStreamEvent;
import com.anthropic.models.messages.Tool;
import com.fasterxml.jackson.databind.ObjectMapper;

void main() throws Exception {
    AnthropicClient client = AnthropicOkHttpClient.fromEnv();
    ObjectMapper objectMapper = new ObjectMapper();

    Tool weatherTool = Tool.builder()
            .name("get_weather")
            .description("Get current weather for a city")
            .eagerInputStreaming(true)
            .inputSchema(Tool.InputSchema.builder()
                    .properties(Tool.InputSchema.Properties.builder()
                            .putAdditionalProperty("city", JsonValue.from(Map.of("type", "string")))
                            .build())
                    .addRequired("city")
                    .build())
            .build();

    MessageCreateParams createParams = MessageCreateParams.builder()
            .model(Model.CLAUDE_OPUS_4_7)
            .maxTokens(1024)
            .addTool(weatherTool)
            .addUserMessage("Weather in Paris?")
            .build();

    // Content block index -> accumulated tool input JSON
    Map<Long, StringBuilder> toolInputs = new HashMap<>();

    try (StreamResponse<RawMessageStreamEvent> streamResponse = client.messages().createStreaming(createParams)) {
        var eventIterator = streamResponse.stream().iterator();
        while (eventIterator.hasNext()) {
            RawMessageStreamEvent event = eventIterator.next();
            if (event.isContentBlockStart()) {
                var blockStart = event.asContentBlockStart();
                if (blockStart.contentBlock().isToolUse()) {
                    toolInputs.put(blockStart.index(), new StringBuilder());
                }
            } else if (event.isContentBlockDelta()) {
                var blockDelta = event.asContentBlockDelta();
                if (blockDelta.delta().isInputJson() && toolInputs.containsKey(blockDelta.index())) {
                    toolInputs.get(blockDelta.index()).append(blockDelta.delta().asInputJson().partialJson());
                }
            } else if (event.isContentBlockStop()) {
                var blockStop = event.asContentBlockStop();
                if (toolInputs.containsKey(blockStop.index())) {
                    var parsedInput = objectMapper.readTree(toolInputs.get(blockStop.index()).toString());
                    IO.println("Tool input: " + parsedInput);
                }
            }
        }
    }
}

<?php

use Anthropic\Client;
use Anthropic\Messages\InputJSONDelta;
use Anthropic\Messages\Model;
use Anthropic\Messages\RawContentBlockDeltaEvent;
use Anthropic\Messages\RawContentBlockStartEvent;
use Anthropic\Messages\RawContentBlockStopEvent;
use Anthropic\Messages\ToolUseBlock;

$client = new Client();

// The PHP SDK does not currently provide a stream accumulator for tool input;
// the manual pattern shown here is the supported approach.
$toolInputs = []; // index => accumulated JSON string

$stream = $client->messages->createStream(
    maxTokens: 1024,
    model: Model::CLAUDE_OPUS_4_7,
    tools: [
        [
            'name' => 'get_weather',
            'description' => 'Get current weather for a city',
            'eager_input_streaming' => true,
            'input_schema' => [
                'type' => 'object',
                'properties' => ['city' => ['type' => 'string']],
                'required' => ['city'],
            ],
        ],
    ],
    messages: [['role' => 'user', 'content' => 'Weather in Paris?']],
);

foreach ($stream as $event) {
    if (
        $event instanceof RawContentBlockStartEvent
        && $event->contentBlock instanceof ToolUseBlock
    ) {
        $toolInputs[$event->index] = '';
    } elseif (
        $event instanceof RawContentBlockDeltaEvent
        && $event->delta instanceof InputJSONDelta
    ) {
        $toolInputs[$event->index] .= $event->delta->partialJSON;
    } elseif (
        $event instanceof RawContentBlockStopEvent
        && isset($toolInputs[$event->index])
    ) {
        $parsed = json_decode($toolInputs[$event->index], associative: true, flags: JSON_THROW_ON_ERROR);
        echo "Tool input: " . json_encode($parsed) . "\n";
    }
}

require "anthropic"
require "json"

client = Anthropic::Client.new

tool_inputs = {} # index -> accumulated JSON string

stream = client.messages.stream_raw(
  model: Anthropic::Models::Model::CLAUDE_OPUS_4_7,
  max_tokens: 1024,
  tools: [
    {
      name: "get_weather",
      description: "Get current weather for a city",
      eager_input_streaming: true,
      input_schema: {
        type: "object",
        properties: {city: {type: "string"}},
        required: ["city"]
      }
    }
  ],
  messages: [{role: "user", content: "Weather in Paris?"}]
)

stream.each do |event|
  case event
  when Anthropic::Models::RawContentBlockStartEvent
    tool_inputs[event.index] = +"" if event.content_block.type == :tool_use
  when Anthropic::Models::RawContentBlockDeltaEvent
    if event.delta.is_a?(Anthropic::Models::InputJSONDelta)
      tool_inputs[event.index] << event.delta.partial_json
    end
  when Anthropic::Models::RawContentBlockStopEvent
    if tool_inputs.key?(event.index)
      parsed = JSON.parse(tool_inputs[event.index])
      puts "Tool input: #{parsed}"
    end
  end
end

Tip

Reach for the manual pattern when you need to react to partial input before the block closes (for example, rendering a progress indicator). Otherwise, prefer your SDK's accumulator helper where the first example on this page uses one.

Handling invalid JSON in tool responses

When using fine-grained tool streaming, you may receive invalid or incomplete JSON from the model. If you need to pass this invalid JSON back to the model in an error response block, you may wrap it in a JSON object to ensure proper handling (with a reasonable key). For example:

{
  "INVALID_JSON": "<your invalid json string>"
}

This approach helps the model understand that the content is invalid JSON while preserving the original malformed data for debugging purposes.

Note

When wrapping invalid JSON, make sure to properly escape any quotes or special characters in the invalid JSON string to maintain valid JSON structure in the wrapper object.

Next steps

Streaming messages

Full reference for server-sent events and stream event types.

Handle tool calls

Execute tools and return results in the required message format.

Tool reference

Full directory of Anthropic-schema tools and their version strings.