ComfyUI – v0.3.69

ComfyUI v0.3.69 is live! 🎉

• New `LatentUpscale` node – Upscale in latent space before decoding for sharper results + faster renders.
• Smarter memory handling – Fewer crashes on big batches; VRAM spikes? Not today.
• SDXL Refiner flow fixed – Seamless transitions between base and refiner—no more weird detail jumps.
• Custom nodes reload properly – Finally! No more restarting ComfyUI after editing your favorite custom node.
• WebAPI polish – Better compatibility with external tools and automation scripts.

Perfect for high-res SDXL wizards and node-based tinkerers. Update now and keep those latents crisp! 💫

🔗 View Release

Ollama – v0.12.11

🚀 Ollama v0.12.11 just dropped — and it’s a quiet gem for the detail-oriented folks!

The big win? `logprob` now includes byte-level data 🎯

No more guessing which bytes map to your tokens. Whether you’re debugging multilingual text, tracking tokenization edge cases, or building precision prompt tools — you now see exactly what’s happening at the byte level.

Perfect for:

• Prompt engineers wrestling with weird token splits
• Researchers analyzing model confidence down to the byte
• Devs building LLM debuggers or token analyzers

No UI fluff, no breaking changes — just pure, nerdy utility.

Upgrade with `ollama pull` and start seeing the hidden layers beneath your prompts. 💡

🔗 View Release

Lemonade – 9.0.2: C++: General Availability (#549)

🚨 Lemonade v9.0.2 just dropped — and C++ is officially GA! 🎉

No more beta labels. The C++ server is now production-ready, faster, leaner, and built for real-world LLM serving. Here’s the breakdown:

• ✅ C++ is now the future — Python NSIS installer and dev tools are gone. Focus fully on C++ for peak performance.
• 🚫 Python server is deprecated — start migrating your workflows now.
• 🔧 Fixed `make_http_request()` bugs + default host = localhost (no more weird network issues).
• 📚 Docs overhauled — cleaner setup, less confusion.
• 💥 Removed all `lemonade-server-dev` clutter — clean, production-ready codebase.
• 📈 All version numbers bumped for clarity.

If you’re running LLMs locally on Ryzen AI or Radeon GPUs — this is your moment. Drop the Python server, go C++, and unlock low-latency AI at scale. 🚀

Time to sip something cold… and run LLMs faster than ever.

🔗 View Release

Ollama – v0.12.11-rc1

Ollama v0.12.11-rc1 is here — and Windows GPU fans, rejoice! 🎉

Vulkan support is officially back on track. No more “Vulkan not found” errors — your RTX, RX, or Arc cards can now accelerate LLM inference without a hitch.

This is a quiet patch with huge impact: if you’ve been stuck on CPU-only runs, it’s time to fire up your GPU again.

⚠️ Still a release candidate — but if you’re on Windows and craving faster generations, this is the one to try.

Pro tip: Update your Vulkan drivers first! Ollama’s fixed its end — now let your hardware do the heavy lifting. 🚀

🔗 View Release

Lemonade – v8.2.2

Lemonade v8.2.2 just dropped—and it’s a game-changer for local LLM tinkerers! 🚀

• Vision-Language Models are live 🖼️🧠: Run LLaMA-based VLMs locally—image + text reasoning, no cloud needed.
• Precise device control: `–device` flag now actually works—tune GPU/CPU with zero guesswork.
• Linux stability fixed: No more crashes or phantom DLL deps. CLI’s solid now.
• HF_HUB_CACHE supported: Smarter offline caching for Hugging Face models—perfect if your internet’s spotty.
• Web UI glow-up: Cleaner layout + new enable_thinking toggle to make models pause & reason before replying.
• Real-time stats endpoint: Monitor `prompt_tokens` live—ideal for optimizing prompts and performance.
• FLM Chat Completions patched: No more broken mid-convo responses.

All wrapped in faster inference and cleaner C++ code. If you’re running LLMs on Ryzen AI or Radeon GPUs—this is your must-update beta. 💪

🔗 View Release

Deep-Live-Cam – Version 2.3c is out now!

Deep-Live-Cam v2.3c is live 🎭✨

Fixed that annoying dropdown bug in model & camera selection—no more weird options or blank menus. Now it just works.

Only available via QuickStart for now (Windows/Mac Silicon users, this is your cue!).

CUDA, CoreML, DirectML, OpenVINO—your GPU still gets the spotlight.

Reload, swap faces in real-time, and keep breaking the internet (responsibly). 🚀

🔗 View Release

Tater – Tater v38

Tater v38 just dropped—and it’s got Siri on speed dial 🥔🎙️

Ask your HomePod “What’s trending on Overseerr?” or “Launch MiSTer” and Tater replies like a chill buddy—short, smart, Siri-optimized. No more robotic jargon. Just clean JSON tool calls and natural voice flow.

🧠 Smarter memory: Each device keeps its own chat history (with TTL caps)—your iPhone won’t remember you asked for pineapple pizza at 3am.

💬 Tighter prompts: Less “according to the model,” more “hey buddy, here’s what’s hot.”

🔧 New plugins:

• 🎮 MiSTer Remote — launch your retro games from the couch
• 🔍 Web Search — instant answers, zero tabs open
• 🎥 Overseerr Request & Trending — see what’s trending and add films with your voice

Now your kitchen counter has a potato-powered AI that knows your movie queue and your SNES library. No laptop needed. Just ask.

Check the README for install tips—Docker or Python, you pick!

🔗 View Release

Ollama – v0.12.10

🚀 Ollama v0.12.10 just dropped — quiet, but mighty!

✅ Code signing is back on macOS & Windows — safer installs, trusted binaries, zero sketchy downloads.

🔧 No flashy new models or UI glitz… just clean, solid housekeeping for devs who run LLMs in production.

If you care about security, integrity, and smooth local inference — this patch is your unsung hero.

Keep those Llama 3, DeepSeek, and Mistral models humming! 🤖💻

🔗 View Release

Text Generation Webui – v3.17

🔥 Text Generation WebUI v3.17 just dropped — and it’s a beast!

💥 Security upgrade: `torch.load()` now defaults to `weights_only=True` — no more sketchy pickle loads in Training_PRO. Safe mode = activated.

🧩 Hugging Face fix: Pinning `huggingface-hub==0.36.0` to squash those pesky venv install bugs. No more “why won’t this install?!” moments.

📊 Training cleanup: `evaluation_strategy` → `eval_strategy` (thanks, @inyourface34456 — clean code wins again!).

🚀 Backend power-ups:

• 🖼️ Qwen3-VL support landed in llama.cpp — multimodal models now run natively. Text + images? Yes, please.
• 🚀 `exllamav3` updated to v0.0.12 — faster, smoother, less stuttering. Your GPU will thank you.

📦 PORTABLE BUILDS ARE LIVE!

No install. No fuss. Just download, unzip, run:

• 🖥️ NVIDIA: `cuda12.4`
• 🖥️ AMD/Intel: `vulkan`
• 💻 CPU-only: `cpu`
• 🍎 Apple Silicon: `macos-arm64`
• 🍏 Intel Mac: `macos-x86_64`

🔧 Upgrading?

Drop your old `user_data/` folder into the new zip — all models, settings, and custom themes preserved.

Go run Qwen3-VL on your old laptop. Your CPU won’t know what hit it. 🤖💥

🔗 View Release

Lemonade – v8.2.1

🔥 Lemonade v8.2.1 just dropped — and it’s production-ready now.

C++ stack got a full glow-up: tray icons, installers, CLI glitches, Ctrl+C crashes — all DEAD. This isn’t a patch; it’s a rebirth.

🚀 Linux users: `.deb` installer landed! `apt install lemonade` on Ubuntu/Debian — no more compiling from source. Mac support polished too.

🧠 Qwen3 GGUF now has `enable_thinking` — unlock deeper reasoning on complex prompts. Think smarter, not just faster.

⚡ `ryzenai-serve` → `ryzenai-server`. Now with halt & reasoning, cleaner logs, and smoother shutdowns. AMD Ryzen folks — this one’s for you.

📊 Real-time memory tracking! Health endpoint shows context size + VRAM usage. No more guessing how much your model is eating.

🧩 Model Manager rebuilt: faster, cleaner, smarter. FLM updates? No more stale models sneaking in.

💻 Web UI & CLI got polish: cleaner commands, better streaming logs, no more redundant callbacks. Feels like a real product now.

📦 Python 8.2.1 synced too — lm-eval fixes, bigger default context/batch sizes for embeddings.

🧪 Test workflows cleaned up. Fewer false CI failures, more confidence in every release.

This isn’t just an update — it’s Lemonade stepping out of beta and into the big leagues.

Grab it. Break it. Tell us how 😄

🔗 Full changelog

🔗 View Release