Content Hub
Open-source AI tools & models, explained
Hands-on setup guides and honest comparisons for the newest AI video, audio, and image models.
Kimi-Audio 2: There's No v2 — What Kimi-Audio-7B Actually Is
There's no Kimi-Audio 2 yet — Moonshot's latest is open-source Kimi-Audio-7B. What it does, the VRAM it needs, and why audio over 30 seconds stalls.
Read the guide →LongCat-Video: Meituan's Open-Source 13.6B Video Model
Install and run LongCat-Video, Meituan's MIT-licensed 13.6B video model — real GPU/VRAM needs, setup steps, and how it compares to Wan 2.2 and Veo3.
Read the guide →