According to reports, the software company Temporal, which developed “durable execution” technology that enables AI agents and other long-running applications to recover from malfunctions without starting over, is seeking about $500 million at a valuation of at least $12 billion.

If completed at that level, the round would double Temporal’s valuation in six months. It would also underscore a gradual shift in investor attention from only “frontier-model” developers toward infrastructure providers that help AI systems run reliably in production.

The shift is notable for the broader AI market. While model developers continue to attract the largest financing rounds, companies addressing the less visible problems involved in deploying AI at scale are also commanding higher valuations. According to Temporal, OpenAI, ADP, Block, and Yum! Brands are among the customers that have deployed its technology.

A valuation that has quadrupled in a year

Temporal’s valuation has risen sharply over the past year. In the first quarter of 2025, the company raised $146 million at a $1.72 billion valuation, then secured a $105 million secondary deal in October that lifted its valuation to $2.5 billion.

In February, Andreessen Horowitz led a $300 million Series D that valued Temporal at $5 billion post-money, with participation from Lightspeed Venture Partners, Sapphire Ventures, Sequoia Capital, and others.

A $12 billion valuation would place the company at more than four times its level from roughly 18 months ago. Temporal has not confirmed the latest fundraising talks.

Why agent builders are paying for durability

Temporal’s appeal lies in a technical problem with outsized consequences. The company describes its platform as a Durable Execution service that allows applications to “ride out” crashes, automatically retry certain operations, and continue from the point of failure instead of restarting from scratch.

That capability is especially important for AI agents. A simple chatbot prompt can be answered quickly, but an agent may run for hours or days, move through multiple steps, depend on external services, and even change course based on model output. The longer such a process runs, the greater the chance that something will fail.

The technology itself predates the agent boom. Temporal was founded in 2019 by Maxim Fateev and Samar Abbas, the creators of the Cadence workflow engine at Uber, and its core engine remains open source, with the company selling a managed service, Temporal Cloud, built on top of it. That open-source distribution helps explain the scale of the adoption figures the company reports, and it means AI workloads now run alongside the long-running business processes—such as payments, order fulfillment, and infrastructure provisioning—the platform was originally built to coordinate. Public case studies list users including Snap, Coinbase, Netflix, Datadog, and Nubank, extending well beyond the customers named in its funding announcements.

Temporal says demand is already visible in its business metrics. When it announced its February funding round, the company said revenue had grown more than 380% year over year, weekly active usage had increased 350%, and installs had risen 500% to more than 20 million per month. Temporal Cloud had processed 9.1 trillion lifetime action executions, including 1.86 trillion for AI-native companies.

In Temporal’s February 17 announcement, OpenAI Vice President of App Infrastructure Venkat Venkataramani said that “as AI systems become more complex and long-running, durability is as important as performance.”

The cost math behind the infrastructure bet

The investment case becomes clearer when the economics of running AI are considered.

On August 17, Gartner predicted that spending on agentic AI workflows will rise fivefold by 2028. Gartner referred to this as the “Inference Paradox,” meaning that even as the cost per token falls, total AI spending rises sharply because far more tokens are consumed.

Will Sommer, a senior director analyst at Gartner, said product leaders “cannot rely on more efficient token economics to rationalize AI costs.” In that context, reliability is not only a technical issue but also a financial one. Each time a workflow fails and must restart, it creates additional model calls, more tokens, and higher computing costs. Temporal believes organizations deploying AI at scale will be willing to pay to avoid that waste.

Temporal is not alone in the category. Major cloud providers already sell managed orchestration tools such as AWS Step Functions, Azure Durable Functions, and Google Cloud Workflows, and newer entrants including Restate and Inngest are also building around durable execution. Temporal differs from the platform-tied options in that its open-source engine runs across clouds and on customers’ own infrastructure, and it is now applying that pre-AI approach to agentic workloads.

Investor behavior suggests a similar view. Crunchbase reported that global venture funding in the first half of 2026 reached a record $510 billion. More than 70% of all second-quarter funding went to AI startups, with OpenAI and Anthropic alone accounting for $217 billion, or 43% of startup funding in the first half of the year.

A Temporal round at a $12 billion valuation would therefore represent more than another large AI deal. It would point to a broader market trend: investors are backing not only AI models themselves, but also the infrastructure needed to make them reliable, scalable, and profitable.