Anthropic

SideSeat automatically extracts model information, token usage, and costs from Anthropic Claude API calls.

Prerequisites

SideSeat running locally (sideseat)
SDK installed (pip install sideseat / uv add sideseat or npm install @sideseat/sdk)
Anthropic API credentials configured

Usage with Anthropic SDK

Python
JavaScript

from sideseat import SideSeat
SideSeat()

import anthropic

client = anthropic.Anthropic()
message = client.messages.create(
    model="claude-sonnet-4-5-20250929",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}]
)
print(message.content[0].text)

import { init } from '@sideseat/sdk';
init();

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

const client = new Anthropic();
const message = await client.messages.create({
  model: 'claude-sonnet-4-5-20250929',
  max_tokens: 1024,
  messages: [{ role: 'user', content: 'Hello!' }]
});
console.log(message.content[0].text);

Extracted Attributes

Attribute	Source
`gen_ai.system`	`anthropic`
`gen_ai.request.model`	Request model parameter
`gen_ai.response.model`	Response model field
`gen_ai.usage.input_tokens`	`usage.input_tokens`
`gen_ai.usage.output_tokens`	`usage.output_tokens`
`gen_ai.request.max_tokens`	Max tokens parameter

Streaming

Streaming responses are captured:

with client.messages.stream(
    model="claude-sonnet-4-5-20250929",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Tell me a story"}]
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

Tool Use

Claude tool use is fully traced:

tools = [{
    "name": "get_weather",
    "description": "Get weather for a location",
    "input_schema": {
        "type": "object",
        "properties": {
            "location": {"type": "string", "description": "City name"}
        },
        "required": ["location"]
    }
}]

message = client.messages.create(
    model="claude-sonnet-4-5-20250929",
    max_tokens=1024,
    tools=tools,
    messages=[{"role": "user", "content": "What's the weather in Paris?"}]
)

SideSeat captures:

Tool definitions
Tool use requests
Tool results

Vision

Image inputs are captured:

import base64

with open("image.png", "rb") as f:
    image_data = base64.b64encode(f.read()).decode()

message = client.messages.create(
    model="claude-sonnet-4-5-20250929",
    max_tokens=1024,
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {"type": "image", "source": {
                "type": "base64",
                "media_type": "image/png",
                "data": image_data
            }}
        ]
    }]
)

Extended Thinking

Claude’s extended thinking feature is captured:

message = client.messages.create(
    model="claude-sonnet-4-5-20250929",
    max_tokens=16000,
    thinking={
        "type": "enabled",
        "budget_tokens": 10000
    },
    messages=[{"role": "user", "content": "Solve this complex problem..."}]
)

Thinking blocks appear in the message thread.

Next Steps

First Run — get started with SideSeat
Python SDK — SDK reference