Signals Inbox·July 20, 2026·AI Creative Tools

Runway vs Kling AI: who leads professional AI video?

Kling AI now narrowly leads professional AI video through stronger models, growth and commercial scale, but Runway still offers the better production environment for teams doing complicated work.

Q1Why do people keep comparing Runway and Kling AI?

Runway and Kling AI are currently the clearest rivals among independent AI video platforms because both want to control far more than a single text-to-video model. They sell generation, editing, reference controls, team tools, APIs, and increasingly complete ways to produce ads, films, social content, and visual effects.

They came from opposite directions. Runway started with creative software, then built its own foundation models. Kling grew out of Kuaishou’s short-video ecosystem and first won attention as a powerful generator.

These days, their products look increasingly similar. Kling 3.0 brings text, images, audio, video, references, storyboards, generation, and editing into one family. Runway offers its own models alongside Kling 3.0, Google models, ByteDance models, editing tools, workflows, and agents.

That overlap makes the rivalry unusually direct. A professional can begin the same project in either product, build a character from references, create several shots, edit footage, add audio, and connect the process to an API.

Kling wants its own model family to handle more of the job. Runway wants the whole job to stay inside Runway, even when another company supplies the best model for one shot.

Q2Why is it still difficult to say which company is winning?

Kling looks stronger as a business and as a model maker, while Runway still feels more complete once a team starts doing real production work. Those advantages pull in opposite directions, so the answer changes depending on what we measure.

Kling gives us unusually hard numbers for a young AI product. Kuaishou has disclosed revenue, annualized revenue run rate, users, enterprise customers, generated videos, and fresh financing. Runway gives us fewer business numbers but far more detail on how studios, agencies, brands, broadcasters, and developers use the product in real work.

The comparison is also becoming less symmetrical. Runway now sells Kling 3.0 inside its own subscription plans. A customer choosing Kling’s model through Runway simultaneously strengthens Kling’s technology position and Runway’s platform position.

Meanwhile, Runway is expanding into agents, developer infrastructure, general world models, and robotics, while the newly separated Kling business remains more concentrated on visual creation.

So far, Kling has the better model and the stronger business. Runway still offers the better working environment.

Q3Which company is commercially bigger today?

Kling AI is now commercially bigger than Runway on the evidence we can verify, and even a generous assumption for Runway leaves Kling ahead.

Kuaishou reported that Kling’s annualized revenue run rate was nearly $500 million in March, after the business generated more than RMB650 million during the first quarter. Kuaishou’s later financing filing also disclosed approximately RMB1.1 billion of Kling revenue for the full year 2025.

Runway has never confirmed a comparable current figure. The most credible widely reported number was a goal of reaching $300 million in annualized revenue by the end of 2025. That was a target, not a reported result. Even if Runway achieved it in full, Kling would still be roughly 67% larger on this measure.

The valuation gap points in the same direction, although valuations reflect expectations rather than sales. Runway’s latest round valued it at a reported $5.3 billion. Kling’s transaction used a $15 billion pre-money valuation, taking the implied post-money value close to $18 billion. Investors are currently pricing Kling at more than three times Runway’s value.

Runway vs Kling AI commercial scale, July 2026

Q4Which company is growing faster now?

Kling AI is growing far faster than anything Runway has publicly demonstrated lately.

Its annualized revenue run rate more than doubled between December 2025 and March 2026, from $240 million to nearly $500 million. That works out to roughly 28% compounded monthly growth. The metric annualizes one month and can jump around, but the first-quarter total backs up the trend: revenue exceeded RMB650 million and grew by more than 300% year over year.

The customer funnel expanded at almost the same speed. Kling said it had passed 100 million global users and was approaching 50,000 enterprise customers by its second anniversary, up from 60 million creators and 30,000 enterprise customers at the end of 2025. Both counts increased by roughly two-thirds in about half a year.

Runway has shipped plenty: new models, agents, developer products, overseas offices, and media partnerships. Yet we still have no current growth series comparable with Kling’s.

The recent numbers leave little room for debate. Kling is growing faster.

Q5Does Kling AI currently have the better video model?

Kling currently has the better proprietary AI video model. The independent rankings have moved decisively since Gen-4.5 briefly led at launch.

Runway announced Gen-4.5 with 1,247 Elo points and first place on Artificial Analysis in December 2025. On the current no-audio text-to-video leaderboard, Kling 3.0 Pro sits in the top five with 1,245 Elo, while Gen-4.5 no longer appears in that group. Gemini Omni Flash and ByteDance’s Seedance 2.0 now occupy the top two positions.

Kling also puts more into one model. Video 3.0 can generate up to 15 seconds, produce synchronized multilingual speech and sound, follow multi-shot instructions, use several visual references, preserve brand text, and let creators specify shots through a storyboard. Kling later added native 4K generation.

Runway’s Gen-4.5 remains a capable text-to-video and image-to-video model, but Runway spreads many of those functions across Aleph, Act-Two, audio tools, recipes, and third-party models.

Runway can still produce excellent footage, and a blind preference leaderboard tells us nothing about editing, enterprise security, or teamwork. Isolate the generator itself, though, and Kling is ahead.

Runway vs Kling AI model comparison, July 2026

Q6Which platform gives directors more control and improves faster?

Runway still gives directors more control after the first clip, even as Kling has improved its core model faster lately.

Runway lets a team separate performance, appearance, editing, generation, and post-production. Act-Two transfers a human performance to a character. Aleph 2.0 edits existing footage with text and timed keyframes. Scene Builder, multi-shot tools, object removal, reusable workflows, and APIs make it easier to change one part of a project without restarting the entire process.

Kling has closed part of that gap through Video 3.0 Omni and Motion Control. Its model can now combine references, dialogue, audio, camera instructions, shot duration, and editing requests inside one multimodal process.

The Motion Control technical report describes a system for retargeting body, facial, and hand movement while preserving identity, with an acceleration method that increased inference speed by more than tenfold.

The last year of releases makes the split fairly clear. Kling said it completed 26 model iterations during its second year, and those releases produced visible gains in audio, motion, storyboarding, consistency, and resolution.

Runway’s recent work has spread across Aleph 2.0, Agent 2.0, Agent Skills, Runway Dev, new APIs, and outside models. Kling has improved the core generator faster. Runway has expanded the surrounding production software.

For a creator who wants the strongest result from one prompt and one model family, Kling is now the better starting point. For a director who expects to revise performances, replace elements, combine providers, and repeat a process across many deliverables, Runway remains easier to direct.

Q7Which one is closer to a complete professional production system?

Runway remains the more complete professional production system today.

A team gets proprietary generation, third-party models, video editing, performance capture, audio, agents, workflows, teamspaces, storage, APIs, enterprise administration, and reusable production recipes. Kling offers a broad creative studio and developer platform, but it still revolves mainly around Kling’s own models.

Runway’s decision to carry rival models has become especially important. Its paid plans currently advertise access to Kling 3.0, while the API also includes Google, ByteDance, and other providers.

A professional team can choose Gen-4.5 for one shot, Seedance for another, Gemini Omni Flash for a third, then edit and organize the outputs in the same environment. Runway can therefore benefit commercially even when its own generator loses a benchmark.

Runway also has the better evidence on total project cost. Gen-4.5 costs $0.12 per second through Runway’s API. Kling’s official API pricing varies by model and resolution; Motion Control, for example, costs $0.126 per second at 720p and $0.168 at 1080p.

Those tools do different jobs, so comparing their per-second rates tells us little. Completed workflows give us a better measure. Silverside reports average savings of 65% on project cost and 73% on production time with Runway, while architecture firm KPF says it reduced a two-week animation task to a few hours.

These are selected company case studies, but they measure the full project, where the budget and schedule actually move.

Kling may produce the better clip more often now. Runway still gives a team more ways to turn that clip into a finished, repeatable production.

Q8Which company has the stronger professional customers?

Runway still has the stronger body of named professional work, although Kling can now point to major productions of its own.

Runway’s latest relationships include an expanded Lionsgate program for co-developed productions, a portfolio-wide partnership with Bertelsmann covering businesses such as RTL Group and BMG, and a growing set of published projects across advertising, television, architecture, music, and live broadcasting.

Kling’s strongest recent example is unusually concrete for this market. Kuaishou said Kling helped generate hundreds of high-quality shots for House of David, including large environments and complex battle scenes, and contributed virtual scenes and visual effects to the Chinese historical drama The Age of Peace.

House of David also shows why logo counting can mislead us. Runway publishes its own case study about the series, while Kuaishou cites Kling. Independent reporting describes a hybrid stack of roughly 10 to 15 core AI tools across the production.

Both companies can fairly claim involvement. Neither can claim that one television credit proves ownership of the workflow.

Runway stays ahead because its case studies explain who used the product, what they made, and what changed in the production. Kling probably has deeper work hidden inside its business customer base, but it has not shown enough of it yet.

Q9Who has the stronger distribution?

Kling reaches far more creators and business users today. Its second-anniversary figures put the platform above 100 million users, across 224 countries and regions, with nearly 50,000 enterprise customers. Runway does not publish anything remotely comparable.

Kuaishou gives Kling a natural advantage. It already understands short-form video, creator acquisition, subscriptions, advertising demand, and high-frequency visual content. Kling can reach users through web, mobile, desktop, APIs, and the wider Kuaishou ecosystem, then learn which formats and features pull people back.

Scale should still be interpreted carefully. At the end of 2025, Kling reported 600 million generated videos from 60 million creators, or roughly ten videos per creator over the platform’s lifetime.

That average looks more like a huge pool of people trying Kling than a platform where every registered user works regularly. A much smaller professional core is probably doing most of the heavy generation.

The business-account count is more commercially useful than the headline user total, but Kling has not published average spending, renewal rates, or production volume.

Runway reaches fewer people and does it through enterprise sales, agencies, studios, APIs, film communities, and lately a larger international footprint. New offices and partnerships can improve that reach, but they do not erase Kling’s current advantage.

Kling can put each new release in front of tens of millions of users almost immediately.

Q10Which company is more enterprise-ready?

Runway currently looks safer for a large company that must satisfy procurement, security, and legal teams before creators can use the product.

Its enterprise documentation explicitly lists SOC 2 Type II, single sign-on, configurable permissions, full ownership of outputs, and a policy of not training on customer data. Runway also provides custom credits, teamspaces, workspace analytics, onboarding, and priority support.

Kling has an enterprise business, API access, centralized plans, and tens of thousands of corporate customers. We found much less public detail on independent security certifications, data-isolation guarantees, permission systems, or enterprise implementation.

Kling may have more controls behind the sales process, but Runway is much easier to evaluate and approve from what it publishes.

For procurement teams, that visibility can decide whether an impressive pilot becomes an approved company tool.

Q11Which company is solving professional AI video’s hardest problem?

Runway is still slightly closer to solving the hardest professional problem: making generated video consistent, editable, and repeatable across an entire production.

The industry has largely moved past the stage where a beautiful five-second clip was enough. Teams now struggle with character continuity, shot matching, revisions, rights, collaboration, and the awkward handoff between generated footage and conventional post-production.

Kling attacks these problems directly through multi-shot generation, storyboards, subject references, native audio, Motion Control, editing, and native 4K output. Its latest model can produce a more complete scene before the footage leaves Kling, reducing the number of separate tools a creator needs.

Runway approaches the same problem by breaking production into parts. A team can alter a performance with Act-Two, edit footage through Aleph 2.0, call several outside models, create reusable workflows, manage a shared workspace, and connect the sequence to an API.

That modular approach is less elegant than a perfect all-in-one model. Current models still fail often enough, though, that professionals need several ways to repair or replace a shot.

Until one-pass generation becomes reliably editable, Runway’s extra escape routes remain more useful to production teams.

Q12Which company has more money to keep competing?

Kling has an overwhelming capital advantage right now.

The latest Kuaishou filing shows about $2.8 billion already committed by the initial and additional investors, with room for the round to reach $3 billion. Runway’s latest round was $315 million. Kling has therefore secured almost nine times more fresh capital in its latest financing.

The gap becomes larger once we include corporate backing. Kuaishou will retain control of Kling after the restructuring, while outside investors provide a large standalone balance sheet and stronger incentives for a future listing.

The financing can support training runs, inference subsidies, international expansion, enterprise sales, hiring, and aggressive pricing. Kuaishou’s filing says the proceeds are intended for business expansion, operations, working capital, and team development.

Runway has strong backers of its own, including General Atlantic, Nvidia, Adobe Ventures, AMD Ventures, Fidelity, and others. Its $315 million round would look enormous in most creative-software categories.

Here, Runway is competing with a better-capitalized Kling, Google, ByteDance, and other infrastructure-rich groups while also funding agents, developer products, world models, and robotics.

Kling can fund more training runs, subsidize inference for longer, and absorb more failed bets. Runway has far less room for error.

Q13Which advantage will be harder for competitors to copy?

Runway is harder to replace inside an established production process, but Kling’s scale is growing faster today.

Runway can create switching costs through saved assets, custom workflows, APIs, team permissions, training, approvals, and integrations. Once a company has organized several production stages around one platform, switching becomes painful even when another model produces a better individual clip.

Kling’s defense comes from a different place. It has a huge flow of new users, Kuaishou’s video experience, a rapidly improving model family, broad usage, and enough capital to keep prices competitive.

A better model can still lure creators away, so model quality by itself is fragile. Kling’s combination of model quality, distribution, revenue, and financing is much harder to reproduce.

Runway’s multi-model strategy protects it from benchmark swings, but it creates pressure on its proprietary technology. When users can access Kling or Gemini inside Runway, the platform becomes more useful while the Runway model becomes less central.

Runway appears increasingly willing to make that trade because owning the professional workspace may prove more durable than owning the top model for a few months.

Focus also favors Kling right now. Runway presents itself as a real-world intelligence company with Creative, Dev, and Robotics products. Kling is being separated into a well-funded company centered on AI visual creation.

Runway’s broader research could produce a much larger business later. Kling currently has the more concentrated plan for winning video.

Q14Are Runway and Kling AI even leading the wider video market?

The latest rankings show that Google and ByteDance currently set the technical pace ahead of both Runway and Kling.

Gemini Omni Flash ranks first in both text-to-video and image-to-video, with ByteDance’s Seedance 2.0 also ahead of Kling in several categories. Kling 3.0 Pro remains among the strongest current text-to-video models, while Gen-4.5 has fallen behind the newest leaders.

The wider market changes how we should judge Runway. Runway now integrates several outside models rather than insisting that every professional use Gen-4.5. That makes it less likely to dominate the underlying model market, but more likely to remain useful as model leadership keeps changing.

Kling takes more risk because its economics are strongest when creators use Kling models directly, and its brand rises or falls more closely with model quality. The upside is substantial because the company captures more of the value when its models win.

The downside is obvious: Google, ByteDance, or another provider can move the benchmark again.

The wider market weakens Kling’s case only slightly. Kling remains closer to the model leaders, while Runway has chosen to compete as the software layer that sits across them. Neither company has locked up the category.

Q15Who is winning Runway vs Kling AI right now?

Kling AI is now narrowly winning professional AI video overall, while Runway continues to lead the professional workflow layer.

The case has become too strong to dismiss. Kling is larger, growing faster, reaches far more people, and has turned its model progress into a much stronger business.

Runway still wins several of the hardest professional categories. Its software is better organized for multi-step production, it publishes stronger evidence of deep customer workflows, and its enterprise controls are more mature and more visible.

A studio, agency, or brand that expects to combine models and revise work repeatedly may still choose Runway even when Kling generates the better first clip.

Commercial momentum and proprietary model quality now carry more weight than Runway’s workflow lead. Kling wins both by a meaningful margin today. The workflow gap has also shrunk as Kling adds storyboarding, editing, motion control, audio, and 4K generation.

On balance, Kling has moved ahead.

Kling can make the verdict much clearer by showing renewals, average enterprise spending, named long-term deployments, security documentation, and deeper collaboration features.

Runway can reverse it by disclosing commercial growth that approaches Kling’s, regaining a top-tier proprietary model position, and turning its expanding media partnerships into a visible stream of released productions.

Kling is winning more of the market today. Runway still gives complicated production teams the better working setup. That keeps the race close, but it no longer keeps Runway in first place.

Runway vs Kling AI leadership scorecard, July 2026