Thanks! And yeah, that’s fair - I need to make the model and hardware requirements easier to find on the website.
There are quite a few models available across LLMs, image, video, audio, etc. For image generation specifically, FLUX and Stable Diffusion models are supported. Z-Image isn’t supported currently.
Ltx-2 is supported, including on 32gb Macs, although generation can be pretty memory intensive and performance will depend on the settings you use.
Please join my discord if you’d like to chat. https://t.co/wsf0ciTfrg
🚀 Lekh AI Pro v7.0 is out for macOS.
Biggest update yet:
• Complete redesign
• On-device voice cloning (Chatterbox TTS)
• AI Personas
• Advanced model configurator
• UltraSharp v2 upscaling by Kim2091
• Major Kokoro + Qwen3 TTS upgrades
Everything still runs 100% locally on Apple Silicon using MLX + GGUF + JANG.
No cloud. No subscriptions. Your AI stays on your Mac.
#AI #LocalAI #AppleSilicon #macOS #MacApps #BuildInPublic #IndieHackers #MachineLearning #LLM #GenerativeAI #OpenSourceAI #MLX #GGUF #ArtificialIntelligence #TechTwitter #Productivity #AITools #OnDeviceAI #MacOSApps #VoiceAI #TTS #AIArt #ImageUpscaler #AIChat #IndieDev
@PrismML Bonsai 8B 🌱 + Gemma 4 💎 now on iOS.
Run powerful local LLMs right on your iPhone – fast, private, no cloud.
• Switch models mid-chat
• Fully on-device
• No tracking, no API keys
This is what mobile AI should feel like.
#AI#iOS#LLM#OnDeviceAI#Privacy
🚀 Lekh AI iOS v7.0 is now live
⚡ Dreamshaper XL Lightning (low-memory image gen)
🧠 Bonsai 8B + Gemma 4 support
💬 Chat improvements
🛠️ Bug fixes + performance boosts
More power, less memory – all on-device.
#AI#LocalAI#iOSDev#MLX#StableDiffusion
@ACEStep_Music ACE-Step just landed in Lekh AI Pro 🎵
Turn ideas into music – fully on-device.
No cloud. No limits. Just pure creativity.
Type a vibe → generate a track → iterate instantly.
Your AI studio now makes music.
#AI#MusicGen#IndieDev#LocalAI
❗️Lekh AI Pro v5.0 is live❗️
Your Mac can generate music, transcribe 90+ languages, and render math.
Without touching the cloud.
Voice → text.
Text → equations.
AI → music.
All running locally.
No APIs. No tracking.
Just your machine doing the work.
#AI#LocalAI #OnDeviceAI #IndieDev #BuildInPublic #MacApps #MachineLearning #PrivacyFirst #AppleSilicon #AItools #Pakistán
@CestIvan@GithubProjects Thank you so much. I’ve been working on this for a very long time. I’m using LTX-2.5 for video generation and I just finished implementing music generation with Ace-step 1.5 XL. It’s such an amazing model. I will be releasing this update soon.
@fiz_codes I’m building a local on-device ai for iOS and Mac. Currently have chat, vision, audio, podcasts, sound effects, images, image editing, video generation, bench marking, RAG and more https://t.co/6nFGR2nvHe
@Dev_code_04 I’m building a local on-device ai for iOS and Mac. Currently have chat, vision, audio, podcasts, sound effects, images, image editing, video generation, bench marking, RAG and more https://t.co/6nFGR2nvHe
@BuffetDesigns I’m building a local on-device ai for iOS and Mac. Currently have chat, vision, audio, podcasts, sound effects, images, image editing, video generation, bench marking, RAG and more https://t.co/6nFGR2nvHe
@heyblake I’m building a local on-device ai for iOS and Mac. Currently have chat, vision, audio, podcasts, sound effects, images, image editing, video generation, bench marking, RAG and more https://t.co/6nFGR2nvHe
Lekh AI Pro v4.3 is out 🚀
• Custom SD + SDXL safetensor support
• LoRA support
• Background model downloads
• Gemma 4 fixes
• Performance improvements
Run your entire image pipeline locally. No cloud. No API keys.
Your AI shouldn’t need the internet.
I just dropped a full walkthrough of my local AI app:
• Chat with MLX / GGUF / Jang models
• Memory that persists across chats
• Chat with your files
• Image + audio generation
• Local AI server (OpenAI-compatible)
No cloud. No tracking.
Watch: https://t.co/IvEWOq4oLF
#AI #LocalAI #OnDeviceAI #MachineLearning #BuildInPublic
Most AI tools lock you into their model list.
That’s not power. That’s permission.
Now your device runs what you choose:
@jangqai Jang models → broader ecosystem
MLX → optimized for Apple Silicon
GGUF → massive open model support
And with a built-in converter, you’re not stuck waiting for compatibility.
You bring the model.
Your device runs it.
No APIs.
No approvals.
No cloud dependency.
Just full control over your AI stack – locally.
#AI #LocalAI #OnDeviceAI #MLX #GGUF #MachineLearning #IndieDev #BuildInPublic #MacApps #PrivacyFirst