Magica adds a QA manager for test cases and release risk
Tell Magica what you are releasing, and the new worker covers test cases, bug reports, fix checks, and release risk.
Every Magica update in one feed — release notes, features and changes, straight from the source.
11 updates
Tell Magica what you are releasing, and the new worker covers test cases, bug reports, fix checks, and release risk.
Magica’s new AI worker plans podcast episodes, books guests, preps interviews, edits recordings and publishes finished episodes to YouTube for you.
Kling 4.0 on Magica combines characters, objects and audio in one Omni Reference, with 30-second scenes and start, middle and end frames.
Magica says the new model reasons through requests faster than Opus 5 while its agent still handles research, design, and production.
Magica's image workflow now runs on GPT Image 2.5, improving text rendering, detail, and prompt understanding.
Magica now combines GPT Image 2.5 and Seedance 2.5 so characters stay consistent across shots with cinematic motion.
Chain text, image, video and model nodes into a system that reruns automatically instead of rebuilding the pipeline by hand each time.
The manual counterpart to Magica's Workflows, letting users pick a specific image, video, voice or chat model directly.
One prompt now drives a full campaign end to end, with Magica handling content, analysis and marketing tasks without back-and-forth prompting.
Sonnet 5 plans and directs each shot while Nano Banana Lite renders images faster and cheaper than the prior model.
Describe a shot in text and the new "super agent" handles the creative work a crew would normally do.