Track every dollar by feature, team, and customer across OpenAI, Anthropic, Google, Groq, Mistral, Together, Azure OpenAI, and AWS Bedrock. Hard-cap budgets before the API call — not after the bill arrives.
pip install burnlens
burnlens start
# Dashboard at http://127.0.0.1:8420/uiBills tell you the model, not the why. Your invoice says gpt-4o: $4,287. It doesn't say which feature, which team, or which customer burned it. By the time you trace the spike, it's already on next month's card.
Alerts arrive after the damage. A bad deploy, a runaway agent, or one abusive customer can trigger thousands of API calls before any dashboard turns red. You find out when you open the bill — or when your CEO does.
Every provider is a different silo. OpenAI's usage page. Anthropic's console. Azure Cost Management. Bedrock CloudWatch. No unified view, no way to ask "which feature is our biggest AI spend across all providers."
-
Drop-in proxy. Point your SDK's
BASE_URLatlocalhost:8420. Existing code works unchanged. The proxy is designed for low overhead and supports streaming passthrough. -
Tag what matters. Three request headers (
X-BurnLens-Tag-Feature,X-BurnLens-Tag-Team,X-BurnLens-Tag-Customer) attribute any call to any dimension. Tags are stripped before reaching the AI provider — they never leave your machine. -
Cap before you call. Register an API key with a daily dollar limit. At 100%, BurnLens returns
429before the upstream request is made — not after the bill arrives. 50% and 80% thresholds fire Slack or email alerts. -
One dashboard for supported providers. OpenAI, Anthropic, Google, Groq, Mistral, Together, Azure OpenAI, and AWS Bedrock spend in one unified view. Model breakdowns, waste detection, and budget tracking use versioned provider pricing.
import os, openai
os.environ["OPENAI_BASE_URL"] = "http://127.0.0.1:8420/proxy/openai"
client = openai.OpenAI(default_headers={
"X-BurnLens-Tag-Feature": "chat",
"X-BurnLens-Tag-Team": "backend",
"X-BurnLens-Tag-Customer": "acme-corp",
})Tags are stripped before the request reaches OpenAI. They never appear in any API payload.
Coding agents. Cursor, Claude Code, Cline, Windsurf — attribute cost per PR, repo, or developer. Set a hard daily cap per API key so one runaway agent can't blow the team's monthly budget overnight.
Customer-facing AI. Tag each request with a customer ID. See which customers drive the most cost, alert on thresholds, and optionally route to cheaper models.
RAG and agents. Tag retrieval calls, tool calls, and generation separately. See whether your vector search or synthesis step is the cost driver — and whether it justifies the output quality.
Internal tools. Set per-team monthly budgets, get Slack alerts at 80% and 100%, and export monthly records for comparison with provider invoices.
| Provider | Status | Notes |
|---|---|---|
| OpenAI | Stable | All models, streaming, reasoning tokens |
| Anthropic | Stable | All models, streaming, prompt caching tokens |
| Stable | Gemini 1.5–2.5 (+ 3.x previews), requires patch_google(); Gemini 3.1 Pro pricing covers requests up to 200K tokens |
|
| Groq | Beta | OpenAI-compatible: point GROQ_BASE_URL at /proxy/groq |
| Together | Beta | OpenAI-compatible: set client base_url to /proxy/together |
| Mistral | Beta | OpenAI-compatible: set client base_url to /proxy/mistral |
| Azure OpenAI | Beta | Point client azure_endpoint at /proxy/azure; set BURNLENS_AZURE_ENDPOINT to your resource URL |
| AWS Bedrock | Beta | Claude models; Bedrock API key (Authorization: Bearer, no SigV4); set BURNLENS_BEDROCK_REGION; Global cross-region pricing |
Pricing covers current text/chat models for the supported providers, plus
audio-modality tokens (OpenAI *-audio-preview / *-realtime-preview, billed at
their own per-million rate) and arbitrary flat per-unit fees via each model's
optional unit_prices (e.g. per web-search call). Image and video generation are
still out of scope. Gemini 3.1 Pro's higher rate above 200K input tokens is
excluded (flat-rate schema uses the ≤200K rate). Audio rates should be re-checked
against the provider pricing page — they change less often than text but do move.
| BurnLens | Helicone / Langfuse | Vantage / CloudZero | |
|---|---|---|---|
| Open source | ✓ | Partial | ✗ |
| Local-first (prompts stay local) | ✓ | ✗ | ✗ |
| Hard caps before API call | ✓ | ✗ | ✗ |
| Per-customer attribution | ✓ | ✓ | ✗ |
| Multi-cloud (Azure / AWS / GCP) | Partial | Partial | ✓ |
Zero config required — sensible defaults out of the box. Optional burnlens.yaml:
budget_limit_usd: 500.00
budgets:
teams:
backend: 200.00
research: 100.00
customers:
acme-corp: 50.00
alerts:
slack_webhook: https://hooks.slack.com/...burnlens start # proxy + dashboard on :8420
burnlens top # live cost by model (htop-style)
burnlens report # weekly cost summary
burnlens analyze # waste detection report
burnlens export # CSV of last 7 days
burnlens run -- python app.py # auto-tag a process with repo / dev / pr / branch
burnlens key register <name> # label an API key + set a daily cap
burnlens key list # list registered keys with caps
burnlens keys # today's spend per registered key
burnlens scan claude # import Claude Code session costs from disk
burnlens scan cursor # import Cursor IDE session costs from disk
burnlens scan codex # import OpenAI Codex session costs from disk
burnlens scan gemini # import Gemini CLI session costs from diskIssues and PRs welcome. See CONTRIBUTING.md.
git clone https://github.com/sairintechnologycom/burnlens
cd burnlens
pip install -e ".[dev]"
pytest