Anthropic’s $125.2B TPU Leases as Token Costs Halve

In September 2026, the LLM Token Expenditure Index put the average cost of a million AI tokens at 97 cents, down from $2.07 on May 28. Yet AI model maker Anthropic had committed $125.2 billion to five years of TPU leases, while OpenAI offered discounts to customers willing to commit to one to three years of compute capacity.

An hourly rental became a multiyear obligation

In 2018, Google Cloud offered beta Cloud TPUs starting at $6.50 an hour. That price described access to a machine for an hour, not a customer’s obligation to keep buying its output. Eight years later, Anthropic has committed $125.2 billion over five years for TPU leases, while Broadcom has agreed to lend it up to $42 billion through a convertible note that could help finance them. The hourly rate and the lease total are not comparable prices: one quotes access by the hour, the other commits a model provider for years.

Anthropic’s prospectus describes more than $518 billion in expected spending over 10 years with six AI-infrastructure partners, roughly 80% of it non-cancelable or payable regardless of usage. That aggregate figure does not establish the cancellation terms of the $125.2 billion TPU leases. A model provider can reduce the price of its next token; it cannot reduce a fixed lease payment by leaving the chip idle.

Those obligations do not all work alike. Anthropic’s agreement to pay SpaceX up to $84.5 billion for Nvidia-based compute through 2029 was reported as mostly cancellable on 90 days’ notice. The cancellation clause, not the headline value, determines how much demand risk Anthropic must carry.

A discount can buy a demand forecast

OpenAI offers Guaranteed Capacity at a discount through one- to three-year commitments based on spend levels. Earlier competition for startup customers featured token credits, promotions and one-time bonuses. A credit encourages a customer to try an API; a spend commitment gives the provider a longer view of the demand it must serve. A lab with fixed bills can exchange discounted rates and guaranteed throughput for committed spend, shifting some utilization risk to the enterprise buyer. OpenAI’s offer does not establish that its customer commitments have the same cancellation terms as Anthropic’s infrastructure leases.

Sources also say Anthropic has ended discounts, typically around 15%, after customers exhausted purchased-token allowances and required renegotiations. Anthropic has not confirmed that account as company policy. A buyer needs to inspect the price after its committed volume runs out.

The buyer has reason to resist signing too early. If token costs keep falling, a customer that reserves capacity may pay for certainty it did not need. A customer whose workflow stalls without throughput may value the same contract differently. Neither buyer can settle that choice from a discount percentage alone.

Agents make a seat a poor unit of demand

SemiAnalysis estimates that agentic AI accounts for about 90% of Anthropic’s business; sources cited in the same report say two clients generated nearly a quarter of its 2025 revenue. Those are estimates and sourced claims, not a count of the tokens each customer consumed. A coding or autonomous workflow can call a model repeatedly while one employee oversees it, so a software seat does not reliably describe its compute demand.

Anthropic has placed Claude Code CLI and Claude for Microsoft 365 in early access. Buyers testing either need to measure model calls per completed workflow before committing to capacity.

The chip supplier now helps finance the customer

Broadcom’s convertible note ties a chip supplier’s financing decision to Anthropic’s ability to fund leased compute. Elsewhere in the stack, General Compute obtained a $400 million loan described as the first to use inference-specific chips as collateral. Lenders are assigning value to hardware whose use depends on sustained inference demand.

The buyer can replace a model more easily than a signature

An enterprise can put model access behind a backend interface and preserve the option to change providers when prices, policies or available models change. It can also measure a workflow by completed work rather than by cheap tokens. Neither choice cancels a capacity reservation already signed. Procurement must specify what happens when consumption exceeds an allowance, falls short of a minimum, or moves to another model.

The September index measured an average cost of 97 cents per million tokens used. Anthropic’s prospectus counts payments due even when capacity goes unused.