Runway’s enterprise numbers show what Higgsfield’s US$400M must compete against

Runway is making a direct argument about where enterprise AI video competition is heading: away from a pure model-quality race and toward the operating layer around generation. In an August 20 post, the company said its business has more than doubled this year, net revenue retention has risen above 300%, and one Fortune 20 customer expanded Runway usage more than 17 times.

Those numbers are company-reported and not independently audited. But they still matter as a competitive signal because they describe the type of enterprise behavior rivals such as Higgsfield now have to win: deeper usage, more workflow dependence and more spend after the initial deployment.

Key Takeaways

• Runway says its business more than doubled in 2026 and net revenue retention exceeded 300%.
• The company is positioning model routing, IP protection, data controls and agentic production as enterprise moats.
• For Higgsfield, the competitive challenge is moving beyond generation quality toward workflow ownership and cost control.

Table of contents

Jump to each section:

• Runway’s numbers raise the enterprise benchmark
• The moat is shifting from models to orchestration
• Security and ownership are becoming product features
• Agentic production changes the cost argument
• What Higgsfield’s US$400M now has to buy

Runway’s numbers raise the enterprise benchmark

Runway’s clearest message is that enterprise adoption is becoming a retention story, not just an experimentation story. Sean Holcombe, Runway’s chief revenue officer, wrote: “Our business has more than doubled this year.”

Runway says net revenue retention is above 300%, while one Fortune 20 customer expanded usage more than 17x this year. The figures are company-reported in Runway’s August 20 enterprise update.

NRR above 300% would mean existing customers, in aggregate, are spending more than three times what they did at the start of the measurement period after expansions and contractions are accounted for. Runway does not disclose the underlying cohort, revenue base or calculation details in the post, so the claim should be read as a directional signal rather than a comparable audited SaaS metric.

The more useful point is what the company says is driving expansion. Runway argues that large customers are no longer buying access to a single video model. They are buying a production environment that can absorb new models, automate choices and satisfy legal and security requirements.

Higgsfield raises US$400M for enterprise AI video

Higgsfield raises US$400M as it expands its enterprise AI video push.

ContentGrip

The moat is shifting from models to orchestration

Runway says enterprise customers get day-zero access to its own frontier models and selected third-party systems, including tools from other major AI labs. Its proposed advantage is not exclusivity. It is the ability to route work across models without forcing teams to rebuild their workflow every time a better generator appears.

The company also highlights a media model router designed to choose among available models based on the job. That matters because model quality is becoming easier for competitors to match or aggregate. If an enterprise can access several leading generators through one production layer, the durable value moves toward orchestration, permissions, asset management and the cost of switching systems.

That is also where Higgsfield is heading. Its enterprise pitch increasingly combines model access with a broader creation environment for marketing teams. The category is beginning to look less like a contest over who owns the best model and more like a contest over who owns the workflow around many models.

Security and ownership are becoming product features

Runway’s enterprise post puts unusual emphasis on legal and governance features. It says customers receive uncapped IP indemnification, including for third-party models, along with no training on customer data, full ownership of outputs and controls such as zero data retention, SSO and enterprise security standards.

For procurement teams, those promises can be more important than a marginal improvement in visual quality. A brand may tolerate a slightly weaker generation result if the alternative creates uncertainty over training data, output rights or where sensitive assets are stored.

This is one area where enterprise AI platforms can build defensibility without owning every underlying model. The more legal, identity and data controls are embedded into the production environment, the harder it becomes for teams to replace the platform with a collection of standalone tools.

Agentic production changes the cost argument

Runway is also pushing beyond generation into autonomous production. The company says Runway Agent can handle multi-step creative work and claims some brands are increasing creative volume while lowering costs.

The important shift is from cost per generated clip to cost per completed production task. Enterprise buyers do not ultimately care how cheap one video generation is if employees still have to coordinate prompts, models, revisions, approvals and exports manually.

That makes agentic workflow a more consequential battlefield than another benchmark win. If the platform can reduce the human coordination required around generation, it can capture more of the production budget and make the software harder to remove.

Runway also says it offers private model licensing for companies that want deeper control. That pushes the enterprise proposition further away from a standard SaaS subscription and closer to infrastructure that large organizations can integrate into their own production stack.

What Higgsfield’s US$400M now has to buy

ContentGrip recently reported that Higgsfield raised US$400 million at a US$5.4 billion valuation while expanding from creator-focused AI video into enterprise marketing.

Higgsfield reported US$700 million in annualized revenue in August 2026, up from about US$20 million a year earlier. The figures were reported in ContentGrip’s earlier coverage and originated from the Financial Times.

That capital gives Higgsfield room to compete, but Runway’s latest positioning shows what the spending target increasingly looks like. The enterprise moat is not one feature. It is a bundle of model choice, security, indemnification, output ownership, workflow automation, private deployment options and measurable cost reduction.

Higgsfield’s strength is that it has grown rapidly with creators and social video users, which can translate into a large top-of-funnel and a strong product culture around fast visual creation. Runway is trying to make a different case: that enterprise expansion comes from becoming infrastructure around creative work, not just the generator inside it.

For marketing teams evaluating the category, the practical comparison is therefore changing. The question is no longer simply which platform produces the best-looking clip. It is which platform can fit governance requirements, route work intelligently, reduce production overhead and stay useful as the underlying models keep changing.

That is the enterprise benchmark Higgsfield’s new capital now has to meet.