InfraPrism

Anthropic Provider

InfraPrism provides a drop-in replacement for the official Anthropic Python SDK, enabling cost tracking for all Claude models.

Installation

pip install infraprism

The InfraPrism SDK includes the Anthropic SDK as a dependency.

Basic Usage

Replace your Anthropic import with InfraPrism:

# Before
from anthropic import Anthropic
client = Anthropic()

# After
from infraprism import InfraPrismAnthropic
client = InfraPrismAnthropic()

All existing code continues to work:

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude!"}],
)
print(response.content[0].text)

Adding Cost Attribution

Add entity parameters to track costs:

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Explain machine learning"}],
    entity_type="customer",
    entity_id="acme-corp",
    tags={"feature": "ai-assistant"},
)

Supported Models

InfraPrism tracks costs for all Claude models:

ModelInput (per 1M tokens)Output (per 1M tokens)
claude-3-5-sonnet$3.00$15.00
claude-3-5-haiku$0.80$4.00
claude-3-opus$15.00$75.00
claude-3-sonnet$3.00$15.00
claude-3-haiku$0.25$1.25

Prices are updated automatically. Check the dashboard for current rates.

Streaming

Streaming is fully supported:

with client.messages.stream(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Write a poem"}],
    entity_type="team",
    entity_id="creative",
) as stream:
    for text in stream.text_stream:
        print(text, end="")

Token counts are calculated after the stream completes.

Async Support

Use the async client for async/await patterns:

from infraprism import AsyncInfraPrismAnthropic
import asyncio

async def main():
    client = AsyncInfraPrismAnthropic()

    response = await client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=1024,
        messages=[{"role": "user", "content": "Hello!"}],
        entity_type="customer",
        entity_id="acme-corp",
    )
    print(response.content[0].text)

asyncio.run(main())

System Prompts

System prompts work as expected:

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    system="You are a helpful coding assistant.",
    messages=[{"role": "user", "content": "How do I sort a list in Python?"}],
    entity_type="employee",
    entity_id="eng-123",
)

Privacy Note: System prompts are never sent to InfraPrism—only token counts.

Tool Use

Claude’s tool use is fully supported:

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

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    tools=tools,
    messages=[{"role": "user", "content": "What's the weather in Tokyo?"}],
    entity_type="project",
    entity_id="weather-bot",
)

Privacy Note: Tool definitions and results are not tracked—only token counts and costs.

Vision (Image Inputs)

Claude’s vision capabilities are supported:

import base64

# From URL
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "image",
                    "source": {
                        "type": "url",
                        "url": "https://example.com/image.jpg"
                    }
                },
                {"type": "text", "text": "Describe this image"}
            ]
        }
    ],
    entity_type="customer",
    entity_id="acme-corp",
)

# From base64
with open("image.jpg", "rb") as f:
    image_data = base64.standard_b64encode(f.read()).decode("utf-8")

response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "image",
                    "source": {
                        "type": "base64",
                        "media_type": "image/jpeg",
                        "data": image_data
                    }
                },
                {"type": "text", "text": "What's in this image?"}
            ]
        }
    ],
    entity_type="customer",
    entity_id="acme-corp",
)

Privacy Note: Images are never sent to InfraPrism.

Multi-Turn Conversations

Track costs across conversation turns:

messages = []

# First turn
messages.append({"role": "user", "content": "What is Python?"})
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=messages,
    entity_type="customer",
    entity_id="acme-corp",
    session_id="conv-123",  # Track as part of a session
)
messages.append({"role": "assistant", "content": response.content[0].text})

# Second turn
messages.append({"role": "user", "content": "What are its main features?"})
response = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=messages,
    entity_type="customer",
    entity_id="acme-corp",
    session_id="conv-123",
)

See Sessions for more on conversation tracking.

Error Handling

Errors from Anthropic are passed through unchanged:

from anthropic import APIError

try:
    response = client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=1024,
        messages=[{"role": "user", "content": "Hello"}],
        entity_type="customer",
        entity_id="acme-corp",
    )
except APIError as e:
    print(f"Anthropic error: {e}")

Failed requests are not tracked.

Configuration

client = InfraPrismAnthropic(
    api_key="sk-ant-...",                # Anthropic key
    infraprism_api_key="ip-...",         # InfraPrism key
    infraprism_environment="production",
)

Next Steps