Image," released on October 11th, has become quite a hot topic overseas. Within about 10 days of its release, faster versions, derivative models, and even LoRA training procedures have appeared one ...

We have finally entered an era where locally running multimodal search can be achieved with a single model. On October 6, ...

Reflection released the weights for Beam, its first open-weight model, on Hugging Face under an Apache 2.0 license on October ...

Tenstorrent released the open source beta of the Open Chiplet Atlas (OCA) Harness on October 22, 2026, the company announced.

NVIDIA Run:ai open source release ships full Kubernetes GPU orchestration platform under Apache 2.0 with 4 modular components ...

Google's DeepMind has released Gemma 4 under the Apache 2.0 license, enabling powerful local AI capabilities on devices ranging from high-end servers to smartphones.

Google releases EmbeddingGemma 2, a 740M parameter multimodal embedding model under Apache 2.0, built for on-device search on phones and laptops.

Google's DeepMind has released Gemma 4 under the Apache 2.0 license, enabling powerful local AI capabilities on devices ranging from high-end servers to smartphones.

Cursor earned its popularity fairly and honestly, for what it's worth. Originally built as a fork of Visual Studio Code, it folded chat, autocomplete, and multi-file edits into an editor that felt ...

Recursive SI released Flint, a 40-billion-parameter mixture-of-experts coding model, on1. The weights are on Hugging Face at ...

Mistral Voxtral 3 replaces three separate vendor contracts with one voice agent SDK, outperforming Whisper large-v3 on ...

Recursive released Flint, a 40-billion-parameter open-weight coding model built to run offline on-device; the company says it ...