Measure the carbon footprint of your AI
Route your AI traffic through Tailpipe's proxy for automatic carbon tracking, cost attribution, and usage analytics. Minimal setup, no application rewrite.
Privacy First
Enterprise-grade security by design. We capture only metadata — never content.
Prompts NEVER captured
Message content is relayed to your provider and never logged, stored, or retained.
Responses NEVER captured
Model outputs are returned to you and never logged, stored, or retained.
API keys NEVER logged
Your credentials are forwarded securely and never stored.
Only metadata collected
Usage metadata — model name, token counts, latency, status, and request attribution — never your content.
Supported Providers
OpenAI
GPT-4, GPT-3.5, and all OpenAI models
Anthropic / Claude
Claude 4, Claude 3.5, and all Claude models
Google Gemini
Gemini Pro, Gemini Flash, and all Gemini models
Mistral
Mistral Large, Medium, and all Mistral models
LangChain & Bedrock
Via the Python SDK's wrappers and callback handler — available on request
Simple Integration
Point your base URL at Tailpipe's proxy and add the telemetry header. For Claude Code that's two environment variables; for SDKs it's a one-line config change, with no wrapper and no rewrite.
OpenAI
from openai import OpenAI
# Change base_url and add the Tailpipe telemetry header
client = OpenAI(
base_url="https://<your-proxy-url>/openai/v1",
default_headers={"x-tailpipe-api-key": "tp_your_key_here"},
)
# Use as normal — telemetry captured automatically
response = client.chat.completions.create(
model="gpt-4",
messages=[{"role": "user", "content": "Hello!"}]
)Anthropic / Claude
from anthropic import Anthropic
# Change base_url and add the Tailpipe telemetry header
client = Anthropic(
base_url="https://<your-proxy-url>/tenant/<your-tenant-id>",
default_headers={"x-tailpipe-api-key": "tp_your_key_here"},
)
# Use as normal — telemetry captured automatically
message = client.messages.create(
model="claude-opus-5",
max_tokens=1024,
messages=[{"role": "user", "content": "Hello!"}]
)Claude Code
# Claude Code — set two environment variables
export ANTHROPIC_BASE_URL="https://<your-proxy-url>/tenant/<your-tenant-id>"
export ANTHROPIC_CUSTOM_HEADERS="x-tailpipe-api-key: tp_your_key_here"
# Then use Claude Code as normalEnterprise Ready
Self-Hosted Deployment
Prefer to hold your data yourself? The self-hosted stack keeps the telemetry store in your own AWS account, and Tailpipe reads it cross-account, read-only. Available on request.
Credential-Gated Telemetry
The hosted proxy rejects requests without a tenant credential and records the credential's hash on every event it accepts, so usage is attributable to your account.
SCI-AI Compliance
Captures all fields required by the Green Software Foundation SCI-AI specification for AI carbon tracking.
Built to Scale
Serverless architecture on AWS Lambda and Kinesis Firehose. Scales automatically with your AI workload — zero infrastructure to manage.
Start tracking your AI carbon footprint
Get set up in minutes. Point your base URL to Tailpipe — or ask us about the Python SDK.