Spira Maxima
Spira Maxima turns a script into a finished social video
Spira AI says the model bakes a presenter, B-roll, captions and music into one pass, and that its agents can post it to TikTok, Instagram and X.
News
Launches, updates, demos and everything that shipped.
Astana AI Film Festival
Astana AI Film Festival awards $450,000 to a Latvian director
Organisers say 8,067 films from 125 countries made it the largest AI film competition yet, with $1 million split across eight winners.
TwelveLabs
TwelveLabs Pegasus 1.6 reads first-person video at $1.75 an hour
TwelveLabs' Pegasus 1.6 adds first-person footage support for robot training data, at the same per-hour and token rates as the version it replaces.
Prism
Tencent Hunyuan's Prism renders 2K video with sound in one pass
The Fudan and Tencent Hunyuan model is MIT-licensed and open on GitHub, but native 2K output needs four or more 80 GB GPUs.
Kandinsky 6.0
Kandinsky 6.0 Video ships open weights with 44 kHz audio
Kandinsky Lab released Lite (3B) and Pro (29B) free under MIT, with five-second clips and lip-sync that run on a 16 GB GPU.
60 Minutes
60 Minutes aired an AI version of its own correspondent
CBS had Runway build the synthetic studio intro, which the show says took about five minutes, then told viewers no more AI content is coming.
Synthesia
Synthesia launches Sessions, a two-way AI avatar platform
The AI video company's new product lets people hold live conversations with lifelike avatars, starting with sales roleplay practice and interview-style surveys.
Supernal
Supernal's 19-day AI-assisted film took third in the XPRIZE
Real actors filmed on a gray screen for one day, AI environments built around them, and a finished trailer in under three weeks.
Cloudflare
Cloudflare ships Clef, the first models it trained itself
The two open-weight models return a probability instead of a sentence, accept images and video, and are free to download under Apache 2.0.
CineMe
CineMe goes live to turn a script into film predevelopment
The platform from filmmaker Dan Hartley and former Prime Video UK boss Chris Bird turns a script into photo-realistic images and video, then carries the project through budgeting and scheduling.
Microsoft
Microsoft ships a live transcription model at $0.54 an hour
MAI-Transcribe-2-Streaming returns partial text in about 100 milliseconds, and two new voice models join it for real-time agents.
HeyGen
HeyGen's first video model sells from 1 cent a second
HeyGen Video is a fine-tune of MiniMax's H3 model and generates 5 to 15 second clips with their own audio, at half its regular price through October.
Nikon
Nikon re-reviews its microscopy video winner over AI use
The 2026 Small World in Motion winner says AI only colored the footage after imaging; critics point to a SynthID watermark on the clip.
AI video
Bethesda shuts down a viral AI-made Fallout game
The browser game, built with Claude Opus 5.5, drew 3.5 million views before ZeniMax sent a cease-and-desist over the Fallout name.
AI video
Malaysia warns of legal action over a fake AI video of Mahathir
The Malaysian Communications and Multimedia Commission is reviewing a 27-second TikTok clip that mixes real footage of the former prime minister with AI-generated scenes of him and his late wife.
wan
Wan 3.0 tops AI video's blind Elo leaderboard
Google's Gemini Omni Flash 1.1 sits eighth at Elo 1116, as of October 2, 2026, on an updated board that pins Kling 3.0 Pro at 1000.
grok
xAI's Grok Imagine Video 1.5 Lite ships from $0.02 a second
The xAI video model carries per-resolution rates of $0.02 at 480p, $0.03 at 720p and $0.14 at 1080p, the same tiers its flagship model and fal.ai list.
kelly-clarkson
Kelly Clarkson wants a separate chart for AI music
The singer told the Las Culturistas podcast on September 30 that she loves AI but does not think AI acts should chart against humans, and fans pushed back.
luma
Luma's Ad Variants turns one static ad into five sizes
Luma's new Ad Variants tool resizes and translates one approved static ad into five placements and any language, while keeping the logo, headline and CTA intact.
higgsfield
Higgsfield's Genjutsu Restyle redraws videos in 20+ styles
The new mode, model ID higgsfield/genjutsu/restyle/v1.0, launched September 30 and bills per second of source video from $0.318 at 480p to $1.632 at 1080p.