Meta Vibes for AI creators
Vibes is Meta's AI-only short-form feed in Meta AI. Remix clips and cross-post to Reels/Stories. Early models: Midjourney and Black Forest Labs.
8 min read
TikTok Symphony Agent explained
TikTok announced Symphony Agent at Cannes Lions June 2026. July added Seedance 2.0, Referring-to-Video, avatars, Auto Selection, Asset Manager.
9 min read
Why Sora ended, facts only
OpenAI shut the Sora app on 26 Apr 2026 and the API on 24 Sept 2026. Reasons below are OpenAI's help page and named reporting, not guesses.
8 min read
DeepSeek V4.1 Flash Multimodal Beta
DeepSeek opened a short V4.1 Flash beta around 8-10 Sept 2026. Native multimodal architecture, V4 Flash pricing, ~20 concurrent. Not a full launch.
4 min read
K2 Horizon Open Model Fleet
IFM shipped six Apache 2.0 K2 Horizon models (0.9B-375B-A23B) on 3 Sept 2026 on Hugging Face, with vLLM, SGLang, and Ollama support.
4 min read
Meta Muse Personal AI Agent
Meta Muse is a personal agent that does real work. US rollout via muse.ai, apps, WhatsApp; know when not to connect payments.
5 min read
Rumik Silk Indic Expressive TTS
Muga (tone tags) and Mulberry 1.5 (descriptions) power Rumik Silk Indic/Hinglish expressive TTS. Open Mulberry is preparing, not a confirmed HF drop.
5 min read
GPT-6 Astra will not make your videos
Astra does not generate video, image, or audio. It is OpenAI's computer-use model; use Veo, Seedance or Kling for clips.
10 min read
fal's H3 Max: Faster Hailuo, Capped at 768p
fal Research post-trained MiniMax H3 and shipped it on 27 August at $0.08/sec. It is faster and cheaper than base H3 — and it gives up 2K to get there.
6 min read
Firefly audio is a licence, not a mix
Generate Music/Speech/SFX went GA 21 Aug 2026. The pitch is universally licensed tracks after the German Suno ruling, not mix quality.
6 min read
Cartesia Sonic 4 Review 2026: Is 40ms Worth It?
Cartesia Sonic 4 hits ~40ms time-to-first-audio at $0.05 per 1,000 credits. What sub-50ms actually unlocks, the pricing gap vs ElevenLabs, and where it loses.
7 min read
Suno Lost to GEMA: What the July 2026 Ruling Means
Suno lost its first AI-music copyright trial in Germany. What the GEMA ruling actually bans, what's still open, and what it means for EU creators.
8 min read
Udio Paused Downloads: What to Use Instead in 2026
Udio has the cleanest AI music licensing in 2026, but downloads are paused until its co-licensed UMG platform ships. What to generate with meanwhile.
8 min read
Reading FLUX 3's Arena debut vote count
FLUX 3 debuted at #2 on Arena's text-to-video board on roughly 1,300 votes out of 616,845. Why a debut rank is provisional, and when to re-test a new entrant.
8 min read
FLUX 3 makes Black Forest Labs a video lab
Black Forest Labs trained FLUX 3 jointly on image, video and audio with 20s synced sound. Which Flux still-image workflows should absorb it, and which won't.
7 min read
Google Vids adds avatars and footage editing
Google Vids now does personal avatars and prompt-edits real footage, with SynthID on every clip. What always-on marking means for video that runs off-platform.
9 min read
LMArena is now Arena: fix your citations
LMArena became Arena in January, and Artificial Analysis is a separate company. The five-field citation rule that stops a rank going to the wrong board.
7 min read
Meta Pulled Its Instagram AI Feature in 3 Days
An Instagram feature that auto-enrolled every public account lasted 72 hours. Five checks to run before you build a workflow on a brand-new platform feature.
8 min read
Muse Video ranks fifth and nobody can use it
Meta's Muse Video sits fifth on Arena's text-to-video board and has never shipped. A four-question filter that strips unusable models out before you plan.
8 min read
NO FAKES is still a bill, not a law
S.4591 advanced out of Senate Judiciary on 18 June 2026 and has not been enacted. What it would do, and why state law is the operative regime today.
8 min read
Qwen-Image 3.0 shipped with no weights
Qwen-Image 1.0 and 2.0 were Apache-2.0 with same-day reports. 3.0 landed with no weights, no licence, no report and no model card. What to do now.
7 min read
The Ross appeal and AI training fair use
The Third Circuit heard Thomson Reuters v. Ross on 11 June 2026 and has not ruled. Do not treat the appeal as settled fair-use law.
8 min read
There is no Veo 4: Google's real video lineup
There is no Veo 4. Google's video work runs across Veo 3.1, Gemini Omni Flash, Flow and Google Vids. Here is which of those you can actually name in a brief.
8 min read
TikTok Labels Uploads Carrying Content Credentials
TikTok watermarks uploads that arrive with C2PA credentials attached. That makes attaching provenance a labelling decision your toolchain is already making.
7 min read