Features
Everything you need to run AI like a business
Full-stack AI cost intelligence — from raw token counts to boardroom ROI numbers.
Cost Intelligence
Real-time visibility into every token, every model, every provider — in one place.
Real-time token-level cost tracking
See cost accrue by the second across OpenAI, Anthropic, and Google Gemini. Input tokens, output tokens, cached tokens — all priced accurately at current model rates.
Model cost comparison
Side-by-side cost analysis for the same workflow across models. Find out which model is costing 16x more per request and whether the output quality justifies it.
Spend forecasting
Project end-of-month spend based on current trajectory. Get alerted before you blow a budget, not after.
Budget alerts
Set per-workflow or per-team budgets. Slack and webhook alerts fire at 80% and 100% of threshold — configurable per environment.
ROI Measurement
Connect AI spend to business outcomes. Build the case for your next AI budget request.
Use case attribution
Every AI call is tagged with a workflow, use case, team, and feature. Costs and outcomes roll up across any combination of these dimensions.
Outcome tracking
Define the outcomes your AI produces — code reviews, tickets resolved, leads qualified — and track them as events. CogniSpend calculates cost-per-outcome automatically.
Fixed and revenue ROI modes
Two ROI calculation modes: Fixed (value per outcome × count, e.g., $50 saved per ticket auto-triaged) and Revenue (actual revenue attributed from funnel events). Use whichever fits your workflow.
ROI dashboards
Per-use-case ROI scorecards with trend charts. See which workflows are earning their keep and which ones are burning money with nothing to show.
Universal Coverage
Track every AI dollar regardless of how you run models — API, SaaS tools, or your own hardware.
SDK wrappers for every major provider
Zero-dependency TypeScript SDK with wrappers for OpenAI, Anthropic, and Google Gemini. Streaming supported. Auto-batching, retry logic, and never-throw error handling so tracking never breaks your agents.
Transparent proxy mode
Change the base URL on your existing OpenAI-compatible client to the CogniSpend proxy. Zero code changes, full tracking. Works with any framework or language.
Manual cost entries and CSV import
Track GitHub Copilot seats, Cursor subscriptions, and any other SaaS AI tool as fixed costs with team and use-case attribution. Import historical API spend from provider CSVs to baseline before you instrument.
Self-hosted and GPU infrastructure
Run Llama, Mistral, or any open model on your own hardware? Track GPU and cloud compute costs as infrastructure spend, attributed to the same workflows as your API calls.
Team & Governance
Built for organizations, not solo projects. Controls your finance team will actually trust.
Multi-tenant organizations
Separate environments for each team, business unit, or product. Costs stay within org boundaries. Aggregate across orgs for the executive view.
Role-based access control
Owner, Admin, Member, and Viewer roles. Finance gets read-only access to cost data. Engineers manage their own integrations. Nobody sees more than they need to.
Automated insight engine
Eight automated rules monitor your spend continuously: cost spikes, unattributed spend, model cost comparisons, weekend anomalies, top-performing workflows, low-ROI alerts, and missing config warnings — delivered as prioritized insight cards.
Slack and webhook alerts
Route insight notifications and budget alerts to any Slack channel or webhook endpoint. Configure per-org or per-workflow routing so the right team sees the right alert.
Spotlight
Cost Explorer
See exactly where your AI budget goes — broken down by provider, model, and workflow.
Monthly spend by provider
Spotlight
ROI Dashboard
Every workflow gets a scorecard. Know which agents earn their keep.
| Workflow | Cost | Value Generated | ROI |
|---|---|---|---|
| Customer Support Triage | $4,200 | $33,120 | 7.9x |
| Code Review Agent | $12,400 | $50,800 | 4.1x |
| Doc Generation | $1,800 | $11,160 | 6.2x |
| Sales Prospecting | $6,400 | $4,500 | 0.7x |
Spotlight
Two-line SDK Integration
Wrap your existing client. Zero latency overhead. Never throws.
Before
import OpenAI from "openai";
const ai = new OpenAI();
const res = await ai.chat.completions.create({
model: "gpt-4o",
messages: [{ role: "user", content: prompt }],
});After — with CogniSpend
import { wrapOpenAI } from "@cognispend/sdk";
import OpenAI from "openai";
const ai = wrapOpenAI(new OpenAI(), {
workflowId: "code-review",
});
// Same API — now every call is tracked
const res = await ai.chat.completions.create({
model: "gpt-4o",
messages: [{ role: "user", content: prompt }],
});Know what your AI is costing. Know if it's worth it.
Set up in under ten minutes. No infrastructure changes required. Start with your existing OpenAI or Anthropic client.
Free plan available. No credit card required to start.