New models and features in AI image, video, voice and music generation, from Midjourney, Sora, Veo, Suno and more.
Microsoft just partnered with AI music startup Suno to bring text-to-song generation directly into Copilot. Anyone can now describe a musical style or enter lyrics in chat to spin up a full track with vocals and instrumentation—no music theory required. Why it matters: It bakes generative audio into everyday browsing, turning casual text prompts into instant pop songs.
Former Pakistan prime minister Imran Khan's party just used ElevenLabs to clone his voice for a four-minute campaign speech from behind bars. The virtual rally, generated from his shorthand notes and archive clips, pulled over 500,000 views on YouTube and led the Pakistani government to throttle major social platforms nationwide. Why it matters: Synthetic media is now officially a political workaround against state suppression.
Stability AI just released Stable Zero123, an in-house model that turns single 2D images into novel views and 3D meshes. It improves on Zero123-XL with cleaner training data and camera angle conditioning, delivering a 40x training speedup. Single-view generation needs the same VRAM as SD1.5, though full 3D reconstruction recommends 24GB.
Google just launched MusicFX, an experimental tool that whips up custom music tracks from simple text prompts. Powered by MusicLM, the tool bakes DeepMind's SynthID watermarks directly into the audio to flag AI origin. To steer clear of legal trouble, it outright refuses to generate vocals or mimic specific artists. It is rolling out first to users in the US, Australia, New Zealand, and Kenya.
Snapchat just rolled out a batch of generative AI perks for its $3.99/month Plus subscribers, letting them generate and send images directly from text prompts. The update also adds an AI photo extend feature that fills in backgrounds, alongside co-op Dreams selfies with friends. It is another push to monetize generative tools inside casual chat.
Meta just rolled out Imagine with Meta, a standalone web-based image generator powered by its Emu model. U.S. users can now generate four high-resolution images per text prompt for free without jumping into a messaging app. Meta also pledged to add invisible AI watermarks that survive cropping and compression to boost transparency.
AI art platform Leonardo.Ai just raised $31M to scale its generative creative suite for enterprise and pro creators. With 7 million users generating over 700 million images, the platform stands out by prioritizing granular creative control, custom model training, and a real-time sketching canvas.
Leonardo AI just opened up its Live Canvas feature to all users, free and paid alike. The tool turns rough sketches into finished artwork in real time as you brush on the canvas. The team also teased that more creative features will roll out over the next week or two.
Ideogram just rolled out image prompting, letting creators upload reference photos alongside text prompts to steer generations. The visual control feature is currently available to all Plus subscribers.
Amazon just launched its Titan Image Generator for enterprise developers via AWS Bedrock. The model creates studio-quality images, swaps backgrounds, and automatically embeds invisible watermarks that survive cropping and compression. To sweeten the deal for cautious enterprises, Amazon is also offering full copyright indemnity.
AI video startup HeyGen just raised $5.6 million at a $75 million valuation and launched a near-instant custom avatar generator. Users can now generate a realistic digital twin from smartphone footage in five minutes instead of waiting days for studio production. HeyGen's annual recurring revenue has already jumped to $18 million.
Stability AI just introduced SDXL Turbo, a text-to-image model that generates 512x512 images in just 207ms on an A100 GPU. Using Adversarial Diffusion Distillation, it reduces inference steps from 50 to just one while matching multi-step quality. Why it matters: It brings text-to-image generation into true real-time speeds, though commercial use is not yet permitted.
Pika just raised $55 million in a Lightspeed-led round while rolling out early access for Pika 1.0. The revamped suite goes beyond simple prompt-to-video, adding tools to change a subject's clothing, insert new characters, and expand canvas ratios. It signals that generative video is shifting rapidly from raw generation to precise scene-level editing.
Runway just rolled out Motion Brush for its Gen-2 video generator, letting you paint directly over any subject to command its exact motion and direction. The update also adds curated style presets, fractional camera controls in Director Mode, and sharper base image models.
Stability AI just released Stable Video Diffusion, its first foundation model for generative video. Built on Stable Diffusion, the release includes two image-to-video models capable of generating 14 and 25 frames, with weights on Hugging Face and code on GitHub. It is strictly for research preview right now, but it brings open-source AI into the video race.
Microsoft just clamped down on its Bing-hosted AI image generator after a viral trend of creating Disney-Pixar-style dog movie posters blew up. The tool briefly blocked the term "Disney" outright before switching to scrambling generated logos and text, likely responding to obvious trademark infringement risks.
Google DeepMind and YouTube just introduced Lyria, their most advanced music generation model yet. Alongside it comes Dream Track, a Shorts experiment that lets select creators generate 30-second tracks mimicking styles and vocals of artists like Charlie Puth and T-Pain, all watermarked via SynthID. It is a major push into licensed, generative audio with the music industry's blessing.
Deforum just announced its official Discord bot, bringing SDXL-powered visual and video generation directly into chat channels. The team is kicking off beta testing in waves over the coming days. It basically removes the headache of complex local installations for creators wanting Deforum animations.
Meta just introduced Audiobox, an audio generation model that creates speech, sound effects, and ambient soundscapes using natural language prompts. It allows creators to combine voice samples with text prompts to restyle voices across different emotions and environments. The model is currently limited to select academic researchers alongside an interactive demo.
Polycam just released an algorithmic update for its gaussian splatting tool that automatically eliminates messy "floaters" in 3D scenes. The update lets users generate splats directly from an iPhone app or video uploads, improves background reconstruction, and makes all outputs free for commercial use. It makes capturing clean 3D environments as simple as shooting a quick video on your phone.
