Howdy all!
Many thanks to all of you who participated in our recent feature poll! Your input is incredibly valuable in helping us decide where to focus our efforts next.
The results are in, and with a clear majority, you've chosen:
🏆 Regional Prompting in Inference! 🏆
We're making this our top development priority. We're already scoping out the work and you can look forward to seeing the first implementation of it in the v2.16.0 dev build...
2025-09-22 00:36:43 +0000 UTC
View Post
Hi Patrons,
With v2.15 now wrapping up and getting ready for its public release, our attention is turning to what's next. We've started planning the v2.16 development cycle, and we'd love your input on where we should focus our efforts.
There are several directions we could go, and your feedback will directly influence our roadmap for the coming weeks and months. Below is a poll outlining our current top ideas -please let us know what you think is the most crucial addition or fix ...
2025-09-06 22:01:06 +0000 UTC
View Post
Howdy patrons! We’ve just shipped v2.15.0-pre.2 - a release focused mainly on bug fixes and polish as we prepare for the big public launch. While this one is mostly about tightening bolts and squashing bugs, we couldn’t resist adding some shiny new things: the AI Toolkit package, a Manual Install button for extensions, Next/Previous navigation on Civitai results, Negative Rejection Steering (NRS)<...
2025-08-27 03:51:30 +0000 UTC
View Post
Howdy Patrons! Version v2.15.0-pre.1 is here, and it’s one of our biggest pre-releases yet! The star of the show is a complete overhaul of the Civitai Model Details page - no more cramped dialogs. Now you get a full-page, feature-rich hub with rich metadata, sleek image viewing, and even integrated inference controls to tweak settings right where you explore your models. It’s smoother, faster, and built to give you total clarity over your workflow.
...
2025-08-17 05:39:18 +0000 UTC
View Post
Hey everyone, we’ve just rolled out v2.15.0-dev.2 and it’s packed with new goodies! You can now experiment with two fresh packages: FramePack and FramePack Studio. We’ve also beefed up model downloads in the HuggingFace browser so you can use your token right from Settings → Accounts. Dra...
2025-07-12 21:59:20 +0000 UTC
View Post
(...plus a sneak peek at what we’ve been working on!)
Howdy everyone!
We wanted to give you a quick update on why things have been a bit quieter lately. As part of a highly strategic and very official “office consolidation initiative,” our entire East Coast division has been relocated to our West Coast headquarters.
(Translation: there’s two of us, and one of us moved across the country so we could collaborate more ...
2025-06-23 02:21:50 +0000 UTC
View Post
Hey Patrons!
We're excited to bring you Stability Matrix v2.15.0-dev.1, packed with significant improvements, headlined by a feature many of you have requested: You now can select the Python version used for Package installs!
We automatically download and install new selected versions - they remain managed within Stability Matrix, so it won't affect the rest of your system environment. Alongside this, we've supercharged our Python dependency manag...
2025-05-23 23:09:06 +0000 UTC
View Post
Howdy Patrons!
We've got another exciting update for you with v2.14.0 Preview 2, packed full of fantastic new features and important fixes!
We're thrilled to introduce Accelerated Model Discovery (Beta), exclusively for our Insider+ supporters. To enable it, click on the ⚡ icon in the Civitai Browser and click Enable. This new feature supercharges your browsing experience with dramatically faster and smoother access to online model repositories. Additionally, t...
2025-04-21 03:38:15 +0000 UTC
View Post
Hi Patrons!
We’re thrilled to roll out our latest update: v2.14.0-pre.1!
We've introduced Abstract Syntax Tree (AST) parsing for Inference prompts, making prompt handling more robust and laying the groundwork for some exciting future improvements. You'll also love the new hotkeys (Ctrl+Up/Ctrl+Down) that allow quick adjustment of token emphasis right within your prompts!
Plus, check out the new Plasma Noise addon, perfect for changing the mood of your tex...
2025-04-05 22:18:28 +0000 UTC
View Post
Introducing the Prompt Amplifier (Exclusive Beta Access for Patrons!)
For some time, we've been working on a new approach to prompt expansion. This is a first-of-its-kind tool, leveraging an experimental, cutting-edge language model to enhance and expand your prompts for superior image generation results. And now, as a thank you for your continued support, we're excited to give our Patrons exclusive beta access, as an addition to your existing benefits! You can find it ...
2025-03-07 04:41:24 +0000 UTC
View Post
Hey Patrons!
We're back with another development build - v2.14.0 Dev 3! This update brings some major additions, including support for 🎞️ Wan 2.1 Text-to-Video and Image-to-Video projects, a brand new package for AMD GPU users on Windows, and a new combined autocomplete tag source to supercharge your Inference prompting. As always, we've been busy fixing bugs and boosting performance based on your feedback, including issues with model dow...
2025-03-07 03:14:45 +0000 UTC
View Post
Hey Patrons!
The results are in! 🎉 Thank you all for voting in the v2.14.0 Feature Poll—your feedback is invaluable in shaping the direction of development. Here’s the weighted results of what you voted for:
📊 Poll Results:
🥇 LLM Prompt Refining/Enhancement – 41 votes
🥈 Inference Prompt Queue – 27 votes
🥉 HunyuanVideo in Inference – 20 votes
🎖...
2025-02-08 18:40:14 +0000 UTC
View Post
Howdy Patrons!
We’re super excited to announce the release of v2.14.0 Dev 2! This update is packed with awesome new features and improvements to make your experience even better.
Check out the new Align Your Steps scheduler in Inference for quicker generation times while retaining similar quality. Also, you can now use wildcards in Inference prompts—like {blue|green|red}—to randomly pick items for your prompts and add s...
2025-01-30 06:00:13 +0000 UTC
View Post
Howdy Patrons! Hope you all had an awesome holiday season! We also took some time off to recharge and enjoy the festivities, but we're back and more excited than ever to keep on building.
We're gearing up for v2.14.0 with OpenModelDB support and some cool new Inference addons like CFG Rescale and Align Your Steps, but we want your input! Please take a moment to vote in the poll and let us know which new features you'd love to ...
2025-01-23 02:07:36 +0000 UTC
View Post
Hey Patrons,
We're excited to share the first progress into Stability Matrix v2.14 with our latest development build, focused on refining your AI image generation workflows and bringing you more model integrations with OpenModelDb!
Inference now features a Rescale CFG addon and a convenient button to swap width/height on the fly. We’ve also introduced new Ctrl+Tab/Ctrl+Shift+Tab shortcuts for quick tab navigation and added an OpenModelDB tab in the Model Browser to help you find...
2025-01-09 06:52:01 +0000 UTC
View Post
Merry Christmas, Patrons! 🎄
We've got a special holiday update for you, and are excited to announce the integration of the ComfyUI-Zluda package! For our AMD users, this should deliver a significant performance boost compared to ROCm/DirectML. We'd like to optimize the installation compatibility as much as possible, please let us know if you experience any errors or performance issues! (Thanks to <...
2024-12-25 02:16:47 +0000 UTC
View Post
Hi Patrons!
We've worked extensively on profiling and improving the startup and background tasks performance with this release, and have made significant improvements to startup time and responsiveness since the last dev release! 🚀 Over 200 class files have been reworked to use faster code generation architecture for dependency injection, and some hardware info fetching is now cached when possible. The video shows preliminary testing on our machines, actual times may vary for differe...
2024-12-19 07:57:34 +0000 UTC
View Post
Hi everyone! We're hard at work finishing up some polish on 2.13, including some major startup speed improvements (compared to the last dev/preview builds at least!), OpenModelDB, and more.
Last year we added a fun little (optional) holiday feature by adding Santa hats to the Civitai model browser images. This year we are lacking in ideas for a "holiday-themed" feature, and would love to hear some suggestions from you folks! Ideally we don't want to make a big performance impact ...
2024-12-13 21:09:55 +0000 UTC
View Post
Hi Patrons!
We are delighted to announce the release of 2.13.0 Dev 3! This update brings support for SD3.5 in Inference, including selections for up to 3 text encoders. The Text Encoder/CLIP selection in Inference is now accessible via the ⚙️ button next to the model selector. New base SD and text encoder models can be downloaded from the Model Browser's Huggingface tab.
From your feedback we've also added usability improvements in ...
2024-11-20 06:52:46 +0000 UTC
View Post
Hi all! Dev 2 build is here, and includes our initial work on adding the requested Extension Packs feature from our last features poll. There is now also the SimpleSDXL package and FluxGym training package available, contributed by PRs from @NullDev.
Extension Packs are groups of extensions that save to a .json format which can be shared with other installations. You can also specify a specific version for ...
2024-10-31 01:52:56 +0000 UTC
View Post
Hey all! Here are the results (including weighting) of our last features poll. Thank you to everyone who voted and left suggestion comments on Patreon / Discord!
The Extensions Packs feature received the most votes by far, followed by OpenModelDB and Prompt Queue, and Symlink Manager. We'll work on adding most of these features in time but will start with these top options as a priority.
The first exte...
2024-10-31 01:52:05 +0000 UTC
View Post
Hello everyone!
We’re excited to share some potential new features for v2.13.0 and would love to hear your input on which ones to prioritize. While multiple features might make it into the final release, your feedback will help us focus on what’s most important to you.
Please leave your suggestions and priorities in the comments below!
As always, thank you for your support!
2024-10-08 02:33:12 +0000 UTC
View Post
Hi all o/ We're beginning to work on our features backlog for v2.13, this first development build implements the highly requested feature of selecting a custom folder path for where Stability Matrix stores model/checkpoint files (can be outside of the Data Directory). We've also re-enabled InvokeAI model sharing which is rebuilt via API based connections that should be much more compatible with InvokeAI model format requirements.
This update also fixes various reported issues since the ...
2024-10-02 04:26:11 +0000 UTC
View Post
Hi all! This last preview update adds the often requested feature of manually setting python dependencies for a package to override the default settings. Currently some packages have default dependency versions of packages like PyTorch that are not the newest version for compatibility across different platforms and hardware. We understand some users may want to try newer versions for speed / performance improvements, so you can now manually enter pip package versions during install / edit the...
2024-09-13 23:43:52 +0000 UTC
View Post
Hi again o/, we're excited to bring you Preview 2 - introducing numerous UI enhancements for the Model Browser and new type-to-search for many selection boxes (like for Inference Models) across the app. We've also added some more comprehensive fixes for current and future potential errors from changes to CivitAI APIs.
On the Accounts page and on lykos.ai, you now have the option to register or sign in with a Google account. An op...
2024-08-20 08:07:51 +0000 UTC
View Post
Hey all, apologies for the late post on this update, mostly fixes and some compatibility improvements on macOS. The last build (6afe2ef) for this version should have been available on auto-update and the website since around last week. We were planning to make some more changes to this version but with the amount of changes we moved them to Preview 2 instead - which should be released soon after this.
As always thank you for your continued support! 2024-08-20 05:38:44 +0000 UTC
View Post
Hey all, we're excited to share the latest updates in our ongoing development, with a special focus on the Checkpoint Manager and Flux support in Inference. The Checkpoint Manager now includes a Metadata Editor, along with new features like drag-and-drop enhancements and file size sorting, giving you greater control and flexibility in managing your models. Additionally, we’ve added Flux support in Inference.
We’re also thrilled to announce...
2024-08-09 04:41:09 +0000 UTC
View Post
Hello everyone!
We're excited to share the latest updates in v2.12.0-dev.2! We've introduced several new features, including a Face Detailer module for our built-in Inference UI. You can find it by clicking the Plus (+) button in the Steps section underneath the Sampler options. The necessary models & extensions will be offered for download if you do not already have them.
This interface may not be final as we adjust things based on your feedback. Please feel free to suggest i...
2024-07-28 21:49:37 +0000 UTC
View Post
Hey all, thank you for everyone who voted in the 2.12 Features Suggestions poll! As shown in the weighted results chart it seems the Model Browser performance improvements, ADetailer for Inference, and mashb1tt Fooocus fork package were the top requested features. This initial development release will first include the mashb1t Fooocus fork package feature, while we're working on the other 2 fea...
2024-07-22 07:58:03 +0000 UTC
View Post
Hey all! We have a few ideas for new features in 2.12.0, though we wanted to get your feedback on which we should prioritize. More than one, or all of the features may make it in the final release, but we wanted to hear your priorities to determine which to focus on.
Feel free to leave suggestions in the comments as well!
Thank you!
2024-07-09 00:18:22 +0000 UTC
View Post
