Client-side attribution
Wrap supported LLM clients with a single line so each call can be attributed to a customer without changing the vendor routing path.
TokenOps
TokenOps is a unit-economics platform for AI companies that attributes LLM spend by customer, reconciles vendor invoices, and shows per-customer gross margin. SDK wrappers, MCP tools, and workflow tools included.
Overview
TokenOps is a unit-economics platform for AI companies that turns LLM usage into per-customer cost and gross-margin data. It captures events at the client layer, reconciles those events against monthly vendor invoices, and shows how spend maps to the customers who incurred it.
The product is positioned for teams that need to answer profitability questions from actual usage rather than spreadsheet exports. It supports one-line wrappers for major LLM providers, exposes the data through a dashboard, API, and MCP tools, and is offered through an early-access flow with a free trial on real telemetry.
Source examples on the page show customer-level cost, revenue, and margin views, plus workflows for building per-customer P&Ls, reconciling invoices, detecting anomalies, and simulating price changes. The workflow framing suggests TokenOps is aimed at finance, operations, and engineering teams that need recurring visibility into LLM unit economics.
Core capabilities
Invoice reconciliation
Capture events, price them against tokens, and reconcile monthly vendor invoices so effective cost can be compared with the invoice total.
Per-customer gross margin
Break down spend, revenue, and margin by customer so teams can see which accounts are profitable and where loss-making usage is concentrated.
MCP access for finance queries
Expose financial data through a 30-tool MCP catalog plus four composite workflow tools for use in Claude Desktop, Cursor, Codex, or other MCP-aware clients.
Structured finance workflows
Use composite workflows for per-customer P&L, invoice reconciliation, anomaly detection, and price-change forecasting.
Multi-SDK and multi-vendor support
Support TypeScript and Python SDK usage, including edge runtimes, with wrappers for Anthropic, OpenAI, Amazon Bedrock, Google AI, Vercel AI, and Azure OpenAI.
Common use cases
Customer profitability reporting
Track which customers are profitable by joining captured LLM usage to customer revenue and margin data for a given period.
Vendor invoice reconciliation
Compare invoice totals with captured events to understand effective rate, drift, and savings across vendors and billing periods.
Usage anomaly detection
Watch per-customer cost run-rate and detect unusual spikes before the monthly invoice arrives.
Agent-assisted finance queries
Use MCP-aware assistants to answer finance questions such as customer margin, loss-making accounts, or cost breakdowns in plain English.
Pricing and margin simulations
Model the impact of new seat-based or usage-based pricing against recent usage to see margin effects and potential churn risk.
Pros and Cons
Pros
- Attributes LLM spend by customer instead of only at the workspace or vendor level.
- Reconciles vendor invoices against captured events to surface effective rate and drift.
- Provides both SDK wrappers and MCP tools, giving teams and agents multiple ways to query the same data.
- Supports several major LLM providers through one-line wrappers.
- Includes workflow tools for P&L, reconciliation, anomaly detection, and price-change analysis.
Cons
- Pricing is not published on the page beyond a trial and plan-sizing note.
- The page emphasizes early access, so availability may still be limited.
FAQ
What does TokenOps do?
TokenOps captures LLM events at the client layer, attributes spend by customer, reconciles vendor invoices against those events, and exposes the results through the dashboard, API, and MCP tools.
Which LLM providers does it support?
The source shows one-line wrappers for Anthropic, OpenAI, Amazon Bedrock, Google AI, Vercel AI, and Azure OpenAI, with support for TypeScript and Python.
How do you set it up?
You can install the SDK, wrap your LLM client, and pass a customerId on each call. The page says the first event can be seen in under five minutes.
Is there pricing information available?
TokenOps says it offers a 30-day free trial on your real telemetry with no card required. After the trial, it says plans are sized to usage, but no pricing details are published on the page.
Quick Facts
- Category
- AI finance / developer tool
- Primary use
- Per-customer LLM cost and gross margin tracking
- Access model
- Early access with a 30-day free trial and no card required
- Integrations
- Anthropic, OpenAI, Amazon Bedrock, Google AI, Vercel AI, Azure OpenAI
- Interfaces
- SDK, dashboard, API, MCP tools
- Domain
- lovie.co
TokenOps Alternatives
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repos and verifies changes with compilers, debuggers, sanitizers, and tests.
PromptScout
PromptScout tracks how ChatGPT, Gemini, Google AI Overviews, and Perplexity mention your brand or competitors, with source analysis and website audits.
SaveMRR
SaveMRR is a Stripe retention tool for SaaS teams to scan churn and MRR leaks, then automate recovery with dunning, cancel-save offers, win-back emails, and onboarding nudges.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads in Firecracker micro-VMs with private networking and SDK, CLI, or MCP control.
Hype
Hype is a web tool for finding trending YouTube topics by category, time range, and scoring mode. Spot emerging ideas and review source videos.
Sleek Analytics
Sleek Analytics is a privacy-friendly web analytics tool with real-time visitor tracking, Core Web Vitals, and revenue attribution.