WebP Cloud Introduces Mode Selection — Rapid (Instant Optimization) and Consistency (Eventual Optimization Mode)

It’s hard to believe, but WebP Cloud has already been running for over two years. As our user base has grown, we’ve gained valuable insight into the diverse needs of different users. With this release, we’re introducing a new concept — Proxy Mode — designed to give you more flexibility in how image requests are handled.

Background

Until now, WebP Cloud has always taken the same approach when handling a user’s first image request:

1. Fetch the original image from the origin server.
2. Process it (strip EXIF metadata, convert formats, crop, watermark, etc.).
3. Return the optimized image to the user.
4. Serve subsequent requests for the same image directly from cache.

This “optimize-first” approach has been our biggest differentiator from services like Cloudflare Polish. (For more details, see our article: WebP Cloud compared with Cloudflare Polish.)

However, after countless conversations with our users, we’ve discovered that their needs typically fall into two categories:

1. API-first users — who treat WebP Cloud as an image optimization API. For them, the key requirement is to get an optimized image on the very first load, minimizing local compute costs.

   ✅ They rely heavily on first-request optimization.

2. CDN-first users — who use WebP Cloud primarily as a website CDN. They care less about whether the first load is optimized, and more about fast responsiveness when publishing new content or onboarding, as well as long-term cache hit rates to reduce origin egress costs.

   ✅ They value speed and efficiency under scale more than immediate optimization.

Introducing Proxy Mode

To better serve both types of users, we’re launching Proxy Mode, which comes with two selectable options:

⚡ Rapid Mode (default)

On the first request, WebP Cloud will convert and optimize the image before returning it.

• Best for API-style use cases.
• Guarantees users always get optimized images.
• This is the current behavior for all existing proxies.

⚡ Consistency Mode (new)

On the first request, WebP Cloud will return the origin image immediately (after lightweight processing like EXIF removal or cropping), while the full optimization happens asynchronously in the background.

• Best for high-concurrency, large-scale traffic scenarios.
• Ensures instant responsiveness on first access.
• Subsequent requests will hit the cache and return the optimized version.

Why Consistency Mode Matters

Consistency mode is designed for performance at scale, offering:

• Instant first render — no delays, better first-screen experience.
• Eventual optimization — initial requests may return the origin image, but once cached, all future requests serve the optimized version.
• Lower egress costs — over time, images are delivered in optimized formats (WebP, AVIF, JXL), significantly reducing bandwidth.
• Resilience under load — in high-traffic scenarios (e.g., new content going viral), it prevents bottlenecks and ensures stable throughput.

👉 In short: if first-request optimization isn’t critical, Consistency mode gives you the best trade-off between speed, stability, and cost efficiency.

Rapid vs. Consistency

Looking Ahead

With Proxy Mode, WebP Cloud now offers two distinct paths:

• Rapid Mode → quality-first, ensuring every image is optimized from the start.
• Consistency Mode → speed-first, eventual optimization, designed for scale.

In the future, we may also introduce differentiated billing to more accurately reflect the resource usage of different request types, while maximizing the benefits of our Always Cache feature. Whether you’re using WebP Cloud as an API or running a high-traffic production site, you’ll now have a mode that fits your needs.

Finally, we’re also exploring partnerships with multiple CDN providers to provide more tailored options — including solutions optimized for mainland China. Stay tuned!