New models and features in AI image, video, voice and music generation, from Midjourney, Sora, Veo, Suno and more.
Meta is cooking up a new feature in WhatsApp that turns your selfies into custom AI stickers. Users upload a few reference photos to Meta AI and can then trigger scenes by typing "@Meta AI imagine me" in any chat—dropping themselves into outer space or a forest. It's currently in beta testing on Android, with full control to delete your reference images at any time.
Suno just launched its official iOS app on the Apple App Store, bringing full text-to-music creation to mobile devices. Users can type in text prompts to generate fully produced songs with vocals and instrumentation in minutes. The launch arrives amid intense legal pressure, as major record labels sue the startup over alleged copyright infringement in its training data.
NBC just announced an AI clone of legendary sportscaster Al Michaels to narrate personalized Paris Olympics recaps on Peacock. Viewers can pick their favorite sports and get greeted by their first name in custom morning highlights, with NBC estimating up to 7 million variants—all vetted by human editors before hitting user feeds.
Toys ‘R’ Us owner WHP Global just premiered what it calls the first brand film created with OpenAI’s video generator, Sora. Creative agency Native Foreign revealed that Sora handled roughly 80% to 85% of the footage, but a dozen artists still had to step in with VFX cleanups. It offers a candid early look at how Hollywood and ad agencies are actually mixing generative AI into production pipelines.
Synthesia just launched Synthesia 2.0, upgrading its AI video platform with personalized avatars shot via webcam or phone and multi-language voice cloning in over 30 tongues. It also previewed next-gen full-body avatars that can talk with their hands later this year. It's making studio-grade corporate video production as simple as typing a doc.
Leonardo AI just released Phoenix, a new flagship image model boasting sharper text rendering, tighter prompt adherence, and a natural-language "Edit with AI" tool. While some legacy features like photoreal mode are temporarily sidelined, its deep creative control turns the platform into a fierce challenger to Ideogram.
Runway just announced Gen-3 Alpha, its next-generation video foundation model trained on an all-new multimodal infrastructure. It delivers major improvements in fidelity, motion, and consistency over Gen-2, enabling fine-grained temporal control and photorealistic character generation. It marks Runway's next big step toward building general world models.
Midjourney just launched Year One, a hardcover coffee table book celebrating its past twelve months of community art and tech evolution. The book documents standout generations from V4, V5, and V6 alongside their prompts, feature milestones, and member profiles. It ships globally from Canada, though initial stock has already sold out.
The Tribeca Festival just announced Sora Shorts, a dedicated program debuting five short films made entirely with OpenAI’s text-to-video model Sora. The selected filmmakers, including Sundance winner Nikyatu Jusu, received early access under guild-compliant terms. It marks the first time Sora-generated films are hitting a major festival screen.
Fable Studio just unveiled Showrunner, a platform that lets viewers generate, customize, and edit animated series with simple text prompts. Users can tweak dialogue, camera angles, and character actions, or let the AI write the next episode on its own. The platform launched a free testing waitlist alongside 10 original shows, though it is currently limited to animation.
AI music generator Udio just dropped udio-130, a new model capable of generating up to two minutes of continuous music from a single prompt. The update significantly improves long-term coherence and song structure without needing manual clip extensions. For now, it's an experimental perk at discounted credit rates for Pro subscribers, with broader rollout coming soon.
Suno just rolled out version 3.5 of its AI music generator, bumping song length to four minutes in a single take and letting you extend tracks by another two minutes. It also adds a sound-to-song feature that turns any audio clip into a full track with text prompts. Pro users get it now, and free tiers follow soon.
Google just introduced VideoFX, a new generative video tool powered by DeepMind's flagship Veo model that features a storyboard mode for scene-by-scene editing. The company also upgraded ImageFX with brush-based image editing and an Imagen 3 waitlist, alongside a DJ Mode beat-mixer in MusicFX.
Runway just opened on-demand streaming for the top 10 finalists of its second annual AI Film Festival, and the biggest takeaway is that human craft still trumps raw generative tech. While every film leaned on AI for backdrops and effects, warped physics frequently broke immersion. Instead, genuine emotional weight came down to strong scripts, raw vocal performances, and traditional cinematography.
Amazon's Audible has quietly added over 40,000 AI-narrated titles since launching its free virtual voice beta for self-published authors. While indie writers love turning ebooks into audiobooks in just 52 minutes, listeners are begging for a filter button, and voice actors warn their careers are actively being squeezed out.
OpenAI just landed its first major music video project with artist Washed Out's new single "The Hardest Part." Director Paul Trillo spent six weeks cherry-picking 55 clips from roughly 700 Sora-generated prompts to pull off sweeping visuals on a budget, though audience reactions have been sharply mixed.
Udio just bumped its track limit to 15 minutes and stretched extension context from 30 seconds to 2 minutes. That extra memory means the model can remember parts from earlier in your song, finally making consistent choruses possible in long mixes.
Ideogram just rolled out a new Pro subscription tier for power creators at $48 a month billed annually. Subscribers get 3,000 fast prompts to churn out up to 12,000 images monthly in under 15 seconds, plus an upcoming CSV bulk upload feature. It supports commercial use, though enterprise copyright indemnification is still missing.
Meshy just launched Meshy-3, rolling out high-poly sculpture generation and complete PBR texture maps straight from text prompts. The release also overhauls the Image-to-3D pipeline and adds a smart healing tool to clean up messy AI artifacts, making generated assets far more practical for game engines.
Shengshu Technology and Tsinghua University just unveiled Vidu, a text-to-video AI capable of generating 16-second clips at 1080p resolution in one click. Built on their U-ViT architecture, it supports multi-camera transitions and simulates real-world lighting. While it doesn't quite match Sora's photorealism yet, its temporal consistency marks a serious step forward for Chinese video models.
