My name got featured in one of Apples WWDC 2025 session videos 👀🍎 Crazy to think my work made it into a global Apple dev session. I’m proud!! Thanks @awnihannun @angeloskath @Apple

Here is a real-life run, end-to-end from training to using the trained LLM in LM Studio by @lmstudio MLX-LoRA-Studio gives you a user friendly, extremely customisable training interface for fine-tuning and training LLMs. Its a MacOS native app, front- and backend. Will be releasing the App in the coming days. GH Repo: github.com/Goekdeniz-Guelmez…

UPDATE 🇪🇺 We passed 10,000 signatures! EU users want Siri AI too. Thanks to everyone who signed and shared. Let’s keep going! siri4eu.com #Siri4EU

You are not ready for mlx-lm-lora by @ActuallyIsaak 😎 Powered by Apple MLX, MLX LM / MLX-LM-LoRA

We should rename it to mlx-omni 🔥

Keep your eye in the repo github.com/Goekdeniz-Guelmez…

Another Sneak Peek of the MLX LoRA Studio App for MacOS

I'm calling it MLX-LoRA-Studio. github.com/Goekdeniz-Guelmez…

What I love about @UnslothAI Unsloth Studio is that it makes training user friendly, the average user gets intimidated by using and writing code, it’s too much hassle, which I understand. But we don’t have a equivalent for Apple Silicon. For now!

Something new will be coming this week too!!!! Sneak peek 👀

Shoutout to MLX-VLM in Apples official WWDC video!! @Prince_Canuma 💪💪💪🔥🔥🔥🔥

It's very cool that Apple shipped a 20B parameter on-device. You can't put 20B parameters in RAM at any reasonable precision. To make it work they are using pretty exotic architecture by today's standards. A small model predicts from the query (or prompt) which experts to load from Nand into RAM. The key distinction from a typical MoE is that you do this once per query and then generate all the tokens with the same experts (instead of switching the experts for every token).

Holy shit, Tim Cooks goodbye speech 🫡😢#WWDC Thank you for everything @tim_cook

Not happing in the EU until the regulations are checked.........

Apple Intelligence!!!!!!!

AHHHHHHH MACOS GOLDEN GATE

It's still so weird knowing this will be Tim's last "Good Morning" #wwdc26

🗓️ Release Week Recap Big week. mlx-audio and mlx-vlm are now among some of the fastest-growing OSS projects. Here’s what we shipped last week. Gemma 4 on Apple Silicon Two awesome releases by our partner @GoogleDeepMind & @googlegemma : > Gemma 4 12B — their new dense, unified multimodal model. It uses an encoder free audio path and simplified vision encoder. > Gemma 4 QAT — quantization-aware training checkpoints, optimized to run locally on consumer GPUs and edge devices, compressing the model while preserving the quality you expect from Gemma 4. On the audio 🎧 side we added support for 15 new TTS, ASR & VAD models, faster long-form transcription, and an expanded OpenAI-compatible audio server. All local on Apple Silicon. Huge thanks to every contributor and my co-maintainer @lllucas. 🙏🏽