Models & Labs
Google Launches Gemini Omni 1.2 Flash and Free Video Gen
Matt Wolfehigh confidence
Why it matters
- →Makes high-quality AI video generation accessible within standard Google Workspace tools.
- →Gemini Omni 1.2 Flash offers a competitive alternative for low-latency API applications.
- →Highlights the shift from standalone AI apps to integrated productivity features.
Google has rolled out Gemini Omni 1.2 Flash, an updated version of its efficient model designed for speed and cost-effectiveness in API usage. Concurrently, Google is expanding its consumer video tool, Google Vids, to include free AI-generated video capabilities. This move integrates generative video directly into the workspace ecosystem, allowing users to create avatar-led videos without additional subscription tiers for basic generation. The release underscores Google's strategy of embedding AI features deeply into productivity suites.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1Google Vids Enhances Video Creation with AI Tools
Google AI Blog · July 16, 2026 · Same story
- 2Google Vids Adds AI Avatars and Gemini Omni
TechCrunch AI · July 16, 2026 · Same story
- 3Google's Gemini 3.5 Pro Model Still Missing
The Rundown AI · July 22, 2026 · Related
- 4Google Introduces Gemini Flash Models
Matt Wolfe · July 24, 2026 · Same story
- 5Google DeepMind Launches Gemini Omni 1.1 Flash
Google DeepMind · August 27, 2026 · Same story
- 6Google's August AI Innovations: Gemini and Beyond
Google AI Blog · September 1, 2026 · Related
- 7Google Launches Gemini 3.8 Flash Model
The Verge AI · September 2, 2026 · Related
- 8Google Releases Gemini 3.8 Flash Model
Matthew Berman · September 3, 2026 · Related
More from Matt Wolfe
© Matt WolfeOpen Sourceimage
Pruna Releases PH3 Image Open-Weight Model
Pruna has released PH3 Image, an open-weight image generation model available on Hugging Face with specific hardware requirements.
Matt Wolfe
© Matt WolfeOpen Sourcevideo
Pika Releases H3 Flex Open-Weight Video Model
Video AI startup Pika has released its H3 Flex model as open-weight, allowing developers to run it on their own infrastructure.
Matt Wolfe
© Matt WolfeOpen Sourcemodels
Kimi K3.1 Pro Open Weights Released by Moonshot AI
Moonshot AI has released the open weights for Kimi K3.1 Pro, a high-performance language model competing with top US labs.
Matt Wolfe
More in Models & Labs
Models & Labsother
llama.cpp b12177 adds CUDA 13 and ROCm 10
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13.5 and ROCm 10.0 alongside existing CUDA 12 builds. AMD GPU users finally get first-class parity with NVIDIA, while Snapdragon mobile chips gain dedicated NPU acceleration paths. Apple Silicon KleidiAI is explicitly disabled here, a notable regression for Mac performance that suggests ongoing tuning issues. The inclusion of OpenVINO and SYCL backends further cements llama.cpp as the universal runtime for inference across nearly every major silicon architecture.
llama.cpp Releases
Models & Labsmodels
llama.cpp adds Kimi K3.1-Pro support
Kimi’s new K3.1-Pro model joins the llama.cpp ecosystem, bringing a distinct architectural twist to local inference. By replacing full-attention MLA layers with DSA (Dynamic Sparse Attention) and linear-attention KDA layers, this release tests how well sparse caching handles hybrid attention patterns. The implementation keeps indexer precision high to avoid quality loss during top-k selection, ensuring the model runs accurately on consumer hardware. This expands the library’s reach into newer Chinese open-weight models that rely on complex memory management.
llama.cpp Releases
Models & Labsother
llama.cpp b12182 release with CUDA 13 and ROCm 10
This release quietly cements llama.cpp as the universal inference runtime by adding default builds for CUDA 13.5 and ROCm 10.0 across Linux and Windows. The inclusion of ROCm 10 means AMD GPU users no longer need custom compilation to access modern acceleration, effectively closing the gap with NVIDIA's ecosystem support. While KleidiAI on Apple Silicon is disabled in this build, the broader platform coverage ensures that local inference remains accessible on virtually any hardware configuration without manual setup.
llama.cpp Releases