TLDR

Fiddl.art now lets you animate any image into a video with a single click. We’ve also added full SVG model support, unified the browse page for images and videos, and introduced the new Seedream3 model. These updates, along with dozens of refinements, make it easier to create, edit, and share dynamic media.

Turn Images Into Video

You can now generate stylized videos from any image on the platform. Whether it’s a portrait or AI-generated artwork, turning it into motion takes just a few clicks. A Create Video button appears on your images, and the backend handles upscaling, default settings, and model compatibility automatically.

This feature works with any image made on Fiddl.art, supports both image-input and non-image-input models, respects aspect ratios, and includes auto-refunds (minus a small fee) if generation fails. It’s designed to give you more creative freedom with minimal friction.

Full SVG Model Support

Our Recraft SVG model now outputs true SVG files. Downloads include properly labeled .svg files, and upscaling is disabled since vectors don’t require it. This makes it simpler to use AI-generated vectors in design and print workflows without conversion hassles.

Browse Page Now Supports Video

The browse page now displays images and videos in a unified gallery. You can filter by media type or browse everything together, making discovery more intuitive. This is especially useful for creators working across multiple formats.

Visuals show a clean, grid-based layout with clear media indicators.

New Image Model: Seedream3

The Seedream 3 model is now available for image creation. It supports more aspect ratios and offers improved style flexibility. Costing 10 points per generation, it integrates seamlessly with existing pricing, model selection, and editing tools like the prompt tuner.

More Predictable Editing

We’ve made several improvements to editing:

  • The edit dialog opens reliably, regardless of image ownership.
  • Prompt edits, aspect ratio limits, and seed settings adapt based on the selected model.
  • Your media tab stays synchronized when switching prompts or loading saved edits.

These changes reduce confusion and help you iterate faster.

Better Video Handling

Video streaming and generation are now more reliable:

  • HD videos stream smoothly with clear error messaging.
  • Failed jobs auto-refund points (minus a service fee).
  • All videos stream securely from S3 using unique access tokens.
  • Aspect ratio and duration settings validate based on the video model.

Every part of the video pipeline has been hardened to minimize failures and improve the user experience.

Technical Cleanup

We consolidated legacy code for better maintainability:

  • Billing logic now uses a single pricing schema.
  • Media types are unified, reducing branching between images and videos.
  • Media gallery components are shared between the browse and create pages, ensuring consistent browsing and filtering.

These under-the-hood improvements make the platform more stable and scalable.

Creator Onboarding Is Accelerating

We’re expanding outreach to grow our creator community:

  • A new Creative Partner Program is in development.
  • Weekly reward challenges run on X (formerly Twitter).
  • Targeted influencer and KOL campaigns are underway.

These efforts aim to build a vibrant ecosystem around Fiddl.art’s tools for generation, sharing, and growth. If you’re new, check out our guide on how to train custom AI models to get started.

Coming Up Next

  • Discoverable custom model library: Browse user-created or fine-tuned models directly in the app.
  • Consistent account attribution: Fixes for missing creator profiles or links on shared content.

Stay tuned for more updates, and explore our recent launch of Sora video support for even more creative options.

Wrapping Up

This update focuses on creative flexibility—turning images into videos, simplifying browsing, and paving the way for deeper collaboration. Try the image-to-video feature from your recent creations, and share what you make. We’re always eager to hear your feedback and ideas.

Follow our development on the Fiddl Trello and join our weekly contests on Fiddl Twitter.

Frequently Asked Questions

How do I turn an image into a video on Fiddl.art?

Click the Create Video button on any image you’ve generated. The system handles upscaling and model compatibility automatically, and you’ll receive a refund if the job fails.

Does Fiddl.art support SVG downloads?

Yes, the Recraft SVG model now outputs true .svg files, ideal for design and print workflows. Upscaling is disabled for SVG downloads.

What is the Seedream3 model best for?

Seedream3 offers improved style flexibility and supports more aspect ratios. It’s great for detailed images and costs 10 points per generation.

Can I browse both images and videos together?

Yes, the updated browse page shows images and videos in a unified gallery with filters to sort by media type.

How does Fiddl.art handle failed video generations?

Failed video jobs auto-refund your points minus a small service fee, ensuring you’re not charged for incomplete work.