1Password launches AI spend tracking for SaaS Manager

1Password announced the public preview of AI Spend and Consumption Management within SaaS Manager, a platform recently recognized as a Leader in the 2026 Gartner Magic Quadrant for SaaS Management Platforms. The tool addresses a mounting crisis: finance leaders across industries are being hit with AI token bills running ten, twenty, even a hundred times over what they forecasted, blowing holes through quarterly budgets.

The problem stems from how AI pricing works. Unlike traditional SaaS applications priced per seat—where headcount tells you the bill—AI is increasingly priced by consumption: every prompt, model call, automated workflow, and autonomous agent adds to the meter. This leaves IT, finance, and AI program leaders unable to answer three critical questions: How much are we spending on AI? Who is driving the cost? How can I make my runway last? Token-based pricing means spend accumulates in real time, not just at renewal. A vendor can quietly shift your default model to a more expensive tier in the middle of a billing cycle, and unless someone notices, every request from that point on costs more with no change in your behavior. A coding agent left unsupervised overnight can burn through a week's worth of tokens on a single task, or on work nobody actually needed. Add in a team ramping a new use case or agentic workflows running unwatched, and a monthly burn rate can double before anyone thinks to check. It is also difficult to get a holistic view of AI spend, since every AI vendor has its own billing model, metrics, and dashboard; getting a credible total requires logging into each portal separately, pulling CSV exports, and maintaining a spreadsheet that becomes stale immediately.

AI Spend and Consumption Management closes these gaps by unifying consumption data from multiple vendors. At launch, it supports Cursor, Anthropic (Claude), and OpenAI (ChatGPT and Platform), with more vendors planned. Teams can set budgets per vendor, configure threshold alerts, and track daily burn rate to see the estimated exhaustion date for prepaid balances. The tool breaks down spend by vendor, team, user, API key, and model so teams can see exactly where consumption is concentrated and find optimization opportunities. Attribution by model and user reveals whether everyone is defaulting to the newest, most expensive model for work a cheaper one could handle—a real financial exposure, since output costs across leading models today vary by roughly 300x for comparable work, from a fraction of a cent to well over a hundred dollars per million tokens. Alerts arrive in Slack and email automatically, so nobody has to remember to check multiple dashboards. SaaS Manager also continuously discovers AI applications employees are using across identity providers, SSO logs, finance systems, device agents, browser extensions, and 1Password vaults; when Flipdish connected SaaS Manager to their identity provider and finance systems, they discovered more than 1,000 applications in under five minutes. AI Spend and Consumption Management is available now in public preview to all SaaS Manager customers at no additional cost, and will be broadly available in Fall 2026.

The root of the AI budgeting crisis is structural: traditional SaaS is priced per seat, so headcount forecasts map to cost, but AI is priced by consumption—every prompt, API call, and agent task adds to the meter in real time. This means a single vendor quietly upgrading a default model mid-cycle, or an unsupervised coding agent looping overnight, can blow through a week's token budget before anyone notices. The damage accumulates before any invoice surfaces it, and visibility is fragmented across separate vendor dashboards, each with its own metrics and billing model. Teams have been operating for two years on an unspoken unlimited plan—experiment freely, figure out ROI later—but as prepaid tranches deplete early and quarterly forecasts get blown, the bill is coming due in a way traditional software never required.

Compounding the visibility problem is governance: the 1Password Access-Trust Gap Report found that over a quarter of knowledge workers use AI tools their employer did not approve, some of which still hit the corporate bill through shared API keys, team subscriptions, or provisioned seats. When an engineer spins up an unapproved tool, IT has no record of it, no insight into what data enters the AI's context, and no way to revoke access when that person moves on. IT sees the SaaS landscape, Finance sees invoices after money is committed, but neither owns a holistic answer to what the company is spending on AI or who is driving the cost.

No comments yet. Be the first to share your thoughts!