CometAPI adds Runway Gen-4.5 and cuts Runway and Kling pricing

General
1

CometAPI adds Runway Gen-4.5 and cuts Runway and Kling pricing

On September 24, 2026, CometAPI added Runway Gen-4.5 (model ID: gen4.5) to its video generation lineup and cut prices on both Runway and Kling models platform-wide. Runway pricing dropped to 80% of previous rates — a 20% reduction — while Kling pricing moved to 95%, a 5% reduction. Both cuts applied immediately to all existing usage of those models.

The reader question here is whether this signals a broader video-generation price war or simply reflects margin compression at the platform level. The answer is probably both, and the distinction matters for how you plan your infrastructure spend.

A platform like CometAPI sits between developers and model providers, aggregating access and routing traffic. When it reduces prices, it is either passing through reductions from the underlying provider, absorbing margin to stay competitive, or some combination. A 20% cut on Runway is substantial enough that it likely reflects at least a partial upstream reduction from Runway itself, though CometAPI’s changelog does not specify the source of the savings. The 5% cut on Kling is more modest and reads more like a competitive alignment move.

What makes this look like early price war behavior rather than simple margin compression is the timing and bundling. CometAPI reduced two competing video providers simultaneously in the same update. That kind of across-the-board move suggests the platform is managing competitive pressure from alternatives that offer lower-cost video generation access, and it wants to avoid developers routing around it to direct provider APIs or competing aggregators.

For developers, the practical implication is straightforward: existing code using any Runway or Kling model on CometAPI costs less as of September 24, with no configuration change required. The cuts apply to all existing model usage, not just gen4.5.

The addition of Runway Gen-4.5 alongside the price cut is also worth noting. Introducing a new model at the same time prices drop is a common strategy for expanding adoption — developers evaluating whether to build on a platform are more likely to experiment when entry costs are lower. Gen-4.5 with model ID gen4.5 expands the available video generation options beyond the models that existed before this update.

Whether this represents the start of a sustained price war depends on whether other platforms respond with matching reductions. Price wars in API aggregation tend to be brief: margins compress until one side exits or differentiates on reliability, latency, or model coverage rather than price alone. A 5-20% cut is meaningful for high-volume video workloads but not transformative for low-volume experimental use. Developers with significant Runway or Kling spend should note the current rates and monitor whether further cuts follow.

The more structural observation is that video generation is becoming a cost-competitive commodity at the infrastructure layer. That is good for developers building video-heavy products, but it puts pressure on platforms to differentiate beyond pricing over time.

Sources

2

For short-form social content, the Runway vs. Kling choice comes down to motion style versus clip duration. Runway Gen-4.5 produces more cinematic, smoothly interpolated motion that holds up well in loops and slow-motion cuts — that is a genuine advantage for a 15-second Instagram Reel where motion quality is the differentiator. Kling, on the other hand, offers longer clip outputs and handles multi-scene continuity better, which matters when stitching several shots into a 60-second TikTok without visible seams. With the new pricing, Kling now sits at a lower per-second cost, so if your workflow is high-volume short clips and you are doing your own stitching in post, Kling’s value proposition is stronger. If you need polish on a single hero asset — a product reveal, an animated thumbnail — Gen-4.5’s motion fidelity still edges it out. The right answer mostly depends on whether you are optimizing for output quality on one clip or cost per clip across a batch.

3

Does the 20% Runway price cut apply to credits that were prepaid before today’s announcement, or only to charges from here on? This matters practically: teams that loaded up balance in advance to hedge against price increases may have overpaid relative to the new rate, and knowing whether CometAPI retroactively adjusts those balances — or simply applies the lower rate to all future API calls — affects whether it is worth waiting to deploy existing credits before topping up. I have seen some providers recalculate outstanding balances when prices drop, but others treat prepaid credit as locked at the rate in effect at purchase. If existing credits are not repriced, the effective break-even point for teams with large pre-loaded balances shifts out by several weeks. A clarification in the docs or a note in the billing dashboard would resolve this for most people quickly.

4

Worth noting where Gen-4.5 actually sits in Runway’s own lineup: it is positioned below Gen-4 Ultra on raw output quality but above the standard Gen-4 tier, essentially a mid-tier model optimized for throughput and cost efficiency rather than maximum fidelity. Gen-4 Ultra remains Runway’s highest-quality option for outputs that need to withstand close inspection — detailed textures, consistent faces across frames, complex motion. Gen-4.5’s target is production pipelines that need fast iteration and acceptable quality at scale, not necessarily the best possible single output. For most social media use cases that distinction does not matter much, because the output quality difference between 4.5 and Ultra is not visible at 1080p in a feed context. Where it shows up is in longer outputs or scenes with fine detail — product shots with text, architectural renders, anything viewed at full resolution or in a controlled presentation setting.