Tater – Tater v57

🚨 Tater v57 — Cerberus Complete is LIVE! 🚨

The AI assistant that talks to any OpenAI-compatible LLM just got a massive brain upgrade — and it’s ready to work smarter, not harder. 🧠⚡

🔥 Meet Cerberus — the 3-Headed Brain**

Tater now runs on a true multi-agent architecture:

🧠 Planner → figures out what needs doing

🛠️ Doer → executes the plan with precision

🔍 Checker → reviews, refines, and validates before sending anything off

✅ Why You’ll Love This:

• 🎯 No more chaotic tool spam — Cerberus picks tools intentionally, not randomly
• 🛡️ Self-recovery when things go sideways (no more stuck loops!)
• 🧹 Cleaner, leaner prompts — perfect for local models & low-resource setups
• 📦 Plugins (vision, weather, HA control, etc.) now integrate consistently across all platforms
• ⏱️ Scheduled & long-running tasks? Way more reliable — no accidental re-scheduling

🗑️ Cleanup & Stability:

• Agent Lab authoring removed (to keep the core tight & fast), but `/agent_lab` still lives as a working dir for logs, downloads, and docs.

This isn’t just an update — it’s the stable foundation for everything coming next: learning, refinement, long-horizon reasoning… the future is cerberus-powered. 🐉

🔗 Grab it now: Tater v57 on GitHub

Let us know how your Cerberus brain behaves! 🤖✨

🔗 View Release

Ollama – v0.16.3-rc0

🚨 Ollama v0.16.3-rc0 is here! 🚨

Big news for Apple Silicon users: Qwen3 model support has landed in `mlxrunner`! 🍏⚡

✅ Qwen3 models now run natively on M1/M2/M3 Macs via Ollama’s MLX backend — no CUDA, no hassle.

🧠 Alibaba’s latest Qwen3 brings stronger multilingual skills and sharper reasoning, making it a serious contender for local LLM workloads.

That’s the headline — this RC is light on changes but heavy on potential 🎯

Stable drop’s coming soon… in the meantime, go test those Qwen3 models! 🧪✨

🔗 View Release

ComfyUI – v0.14.1

🚨 ComfyUI v0.14.1 is out! 🚨

The latest patch is here — and while the GitHub release page is currently having trouble loading (we’re hoping it gets fixed soon 🤞), here’s what we expect based on typical patch releases like this:

🔹 Bug fixes — especially addressing pesky regressions from `v0.14.0`

🔹 UI/UX polish — think smoother node dragging, cleaner error popups, maybe a layout tweak or two

🔹 Performance tweaks — smarter caching, lighter memory footprints, faster node execution

🔹 Dependency updates — safer, more compatible versions of key Python packages under the hood

🔹 Accessibility & locale improvements — better support for international users and screen readers

💡 Bonus: If you’re curious about the exact changes, run:

“`bash

git log v0.14.0..v0.14.1 –oneline

“`

…or keep an eye on the Releases page — fingers crossed it loads soon!

Happy prompting, everyone! 🧠✨

🔗 View Release

ComfyUI – v0.14.0

🚀 ComfyUI v0.14.0 is out — and it’s packing some serious upgrades!

Here’s what’s new in this fresh release 🌟:

• 🧠 Smarter Custom Node Support: Improved loading, error handling, and compatibility — especially for nodes using dynamic imports or `folder_paths`. Fewer crashes, more creativity!
• 🧩 Better Node Discovery & Management: Early groundwork for a built-in node registry + tighter integration with tools like `comfyui-manager`. Soon, installing & updating nodes might feel almost too easy 😎
• 🎨 UI/UX Polish: Smoother zoom/pan, snappier node layout rendering, and refined dark/light theme consistency. Your workflow just got more pleasant.
• 📦 Dependency Fixes: Cleaner handling of optional backends like `xformers`, `bitsandbytes`, and CUDA builds — with smarter fallbacks when things go sideways.
• 🚀 Speed Boosts: Faster graph execution, especially for large or batched workflows. Less waiting, more generating!
• 🛠️ CLI & Headless Mode Love: Enhanced scripting support and API stability — perfect for automation, CI/CD pipelines, or headless servers.

💡 Bonus: If you rely on popular custom nodes (IPAdapter, ControlNet helpers, etc.), this release likely means fewer compatibility headaches and more stable runs.

👉 Check out the full changelog on GitHub or join the Discord for deep-dive threads!

Let’s build something wild with v0.14.0 🎨✨

🔗 View Release

Ollama – v0.16.2: mlxrunner fixes (#14247)

🚨 Ollama v0.16.2 is live! — A focused patch packed with Apple Silicon love and stability wins 🍏⚡

🔹 mlxrunner fixes (issue #14247):

✅ `glm4_moe_lite` now loads smoothly on Apple Silicon via MLX — huge for MoE fans!

✅ Diffusion models (like Stable Diffusion variants) finally play nice 🎨

✅ Logs are much quieter — no more debug spam cluttering your terminal 🧹

✅ `–imagegen` flag now works reliably for image generation workflows

💡 TL;DR: Smoother Apple Silicon experience, better model compatibility (especially GLM & diffusion), and cleaner output — all without flashy new features. Just solid, reliable improvements! 🛠️

Grab the update and keep local LLMing! 🚀

🔗 View Release

Tater – Tater v56

🚨 Tater v56 — Cerberus Upgrade is LIVE! 🚨

The core of Tater just got a massive intelligence overhaul — meet Cerberus, the new 3-phase reasoning engine:

🔹 Plan → Execute → Validate — ensures smarter, step-by-step execution with fewer surprises.

🛡️ Tool Safety Upgraded

• Tools only fire when absolutely intended
• Malformed or accidental calls? Nope. Not today.

✉️ Smarter Messaging

• `send_message` now ignores casual chat — only triggers on clear intent
• “Send it here” ✅, “Hey, send that” ❌ (nope!)

🧠 Context That Actually Makes Sense

• Clean, scoped memory per conversation
• “Do that again?” → Works now
• Topic shifts reset context intelligently (no more weird carryover!)

📊 Behind-the-Curtain Wins

✅ Smoother error handling & retries

✅ Reduced token bloat = lower costs & faster responses

✅ More deterministic behavior (yes, predictable LLMs are a thing now)

🎯 What This Means for You

✔️ Reliable multi-step workflows (finally!)

✔️ Fewer “why did it do that?!” moments

✔️ Natural, fluid follow-ups

✔️ A rock-solid foundation for future agent smarts

🔮 What’s Next?

Cerberus sets the stage for long-horizon tasks, learning, and advanced agent behavior — but v56 is all about stability, safety, and raw, reliable power.

🐶 Cerberus is awake. 🐶🔥

👉 Check the README to upgrade & explore!

🔗 View Release

Heretic – v1.2.0

🚨 Heretic v1.2.0 is live — and it’s packing serious upgrades! 🚨

The team just dropped a massive update to Heretic, the fully automatic censorship-removal tool for LLMs — and this one’s packed with performance boosts, new features, and rock-solid stability fixes. Here’s what’s new:

🔹 Memory & Stability Wins

• `amax_memory` setting to cap RAM usage 🧠
• Smarter iteration logic to avoid getting stuck in low-divergence traps ⚙️
• Magnitude-Preserving Orthogonal Ablation — finer control over refusal suppression 🎯

🔹 LoRA + Quantization Power-Ups

• Brand-new LoRA-based abliteration engine 🪄
• Full 4-bit quantization support — run decensoring on consumer GPUs 🖥️
• Auto-GPU detection at startup (no more manual config headaches!) 🤖

🔹 Resilience & Flexibility

• Run extra optimization trials after the main run finishes 🔄
• Save/resume support for long jobs — no more lost progress! 💾
• Fixed MXFP4 loading issues 🛠️
• Refactored save/load machinery = way more reliable 📦

🔹 Vision-Language Models? Now Supported!

• Broad support for VL models (e.g., LLaVA-style) 📸➡️🧠
• Plus: full type checking, debug tools, slop-reduction configs, and bug fixes 🐞

🌟 Shoutout to @noctrex, @accemlcc, @anrp, and @salmanmkc — welcome to the crew!

🔗 Full changelog: v1.1.0 → v1.2.0

🚀 Grab it now and keep decensoring smarter, faster, and safer!

🔗 View Release

Ollama – v0.16.2-rc0: mlxrunner fixes (#14247)

🚨 Ollama v0.16.2-rc0 is here — and it’s bringing serious Apple Silicon love! 🍏

This release candidate focuses on fixing critical bugs in the `mlxrunner` backend (hello, M1/M2/M3 users 👋), making it way more reliable for cutting-edge models. Here’s what’s new:

🔹 GLM4-MoE-Lite now loads smoothly via `mlxrunner` — no more crashes on M-series!

🔹 Diffusion models (like Stable Diffusion) — finally load without hiccups 🖼️

🔹 Cleaner logs — less noise, more signal (goodbye, endless debug spam 🧹)

🔹 `–imagegen` flag now actually works for image generation workflows ✅

💡 Pro tip: This is a release candidate, so it’s the perfect time to test stability before v0.16.2 drops — especially if you’re into MoE models or local image gen on Mac.

Ready to give it a spin? 🧪 Let us know how it goes!

🔗 View Release

Home Assistant Voice Pe – 26.2.1

🚀 Home Assistant Voice PE v26.2.1 is live!

The latest update (from v25.12.4 → v26.2.1) is all about polish and reliability—perfect for those relying on voice control in their smart homes, especially offline. Here’s what’s improved:

✅ Media playback is smoother & more stable

No more skips or dropouts—your voice-triggered music, alerts, and announcements now play cleanly.

🛠️ TTS timeouts fixed!

Text-to-speech responses now fully render and play—no more truncated or missing voice replies. 🎙️✨

💡 Bonus: The project is now officially sponsored by the Open Home Foundation—a big vote of confidence in its mission for private, local-first voice control. 🏡🔐

78 releases down, more innovation ahead! 🛠️

Check it out if you’re building or expanding a local, privacy-first voice assistant setup. 🎯

🔗 View Release

Ollama – v0.16.1

🚨 Ollama v0.16.1 is live! 🚨

Hey AI tinkerers & local LLM lovers — fresh update incoming! 🔥

What’s new in v0.16.1?

🔹 New model config added: `minimax-m2.5` 🧠

• Looks like a fresh MiniMax model variant (internal/experimental for now — keep an eye out for docs!).
• You can already pull it via `ollama pull minimax-m2.5` if you’re feeling adventurous 🛠️

🔹 Lightweight patch release — no breaking changes, just lean & mean model support upgrades.

📦 Binaries are rolling out for macOS, Windows, and Linux — grab the latest from GitHub or update via your package manager.

👉 v0.16.1 Release Notes

Let us know if you get `minimax-m2.5` running — curious to hear your benchmarks and use cases! 🧪✨

🔗 View Release