OpenMindAI v3.0.1 is a stability and usability update for the OpenMindAI desktop and mobile ecosystem. This release improves automatic model routing, local model detection, mobile chat reliability, model installation, download progress tracking, and hardware-aware model recommendations.
The v3 series introduces a unified AI experience across Windows, macOS, Linux, Android, and iOS with local and offline AI, intelligent model selection, reasoning, coding, vision, image generation, document workflows, PDF/DOCX generation, research tools, and privacy-focused on-device processing.
Version 3.0.1 also includes improved model cards, clearer capability information, better fallback handling, enhanced mobile interfaces, and multiple performance and stability fixes across the platform.
OpenMindAI v3.0.1 is a major stability, usability, and model-system improvement release for the OpenMindAI desktop and mobile ecosystem. This release builds on the foundation introduced with OpenMindAI v3.0.0 and focuses on making local AI more reliable, easier to use, and more consistent across Windows, macOS, Linux, Android, and iOS.
Version 3.0.1 brings important improvements to automatic model routing, local model detection, model installation, download management, mobile chat, hardware-aware recommendations, document workflows, image workflows, application stability, and error handling. It also improves the way OpenMindAI presents models and their capabilities so users can focus on what a model can do rather than dealing with unnecessary technical information.
OpenMindAI v3.0.1 continues the transition toward a unified desktop and mobile AI platform. The same OpenMindAI ecosystem can now support local AI chat, reasoning, coding, vision, document workflows, image-related tasks, research workflows, and intelligent model management across supported platforms.
The v3 architecture is designed around a local-first experience. After the required models and runtimes are installed, supported AI workloads can run directly on the user's device. This allows OpenMindAI to provide offline functionality while reducing dependency on external AI services for tasks that can be handled locally.
OpenMindAI now provides a more consistent experience across desktop and mobile devices. The application architecture has been improved so that model management, routing, chat, downloads, local inference, and capability detection behave more consistently regardless of the platform being used.
Windows, macOS, and Linux remain part of the desktop OpenMindAI ecosystem, while Android and iOS extend the platform to mobile devices. The goal is to provide the same core OpenMindAI experience while allowing each platform to use models and runtimes that match its own hardware capabilities.
The OpenMindAI Mobile application received significant improvements in this release. Local AI chat is now more reliable, installed models are detected more consistently, and the model-routing system has been improved so that mobile users are less likely to encounter errors when starting conversations.
The mobile interface has also been refined to provide a cleaner and more responsive experience. Chat, model management, image workflows, document generation, reasoning, and coding-related features are now better organized for smaller screens.
The application is designed to recommend smaller and more efficient models for mobile devices when appropriate, while more powerful devices can access stronger models depending on available RAM, CPU performance, storage, and supported acceleration.
Automatic model routing is one of the core improvements in OpenMindAI v3.0.1. The system can now more reliably determine which installed model is appropriate for a particular task.
Instead of requiring the user to manually select a model for every conversation, OpenMindAI can evaluate the requested task and match it against the capabilities of available models. This can include general chat, coding, reasoning, deep thinking, vision, document analysis, research, tool-oriented tasks, and agentic workflows.
Manual model selection remains available for users who prefer direct control, but automatic routing makes the default experience simpler and more intelligent.
The routing engine has been improved to better handle installed models, missing models, loading states, capability matching, hardware compatibility, fallback models, and local inference initialization.
Previously, a model could be downloaded successfully but still fail to be selected correctly in certain situations. Version 3.0.1 improves the communication between the model manager, model registry, chat system, and automatic router.
If a preferred model is unavailable, OpenMindAI can now more reliably determine whether another compatible installed model can be used instead.
Local model detection has been improved across the application.
In earlier builds, some models could be present on the device but still appear unavailable inside the chat interface or model manager. Version 3.0.1 improves synchronization between the local filesystem, model registry, model installation state, active model state, and chat routing system.
This reduces unnecessary model installation prompts and improves the reliability of starting local conversations immediately after a model has been downloaded.
A major mobile chat issue has been addressed in this release.
In some cases, users could successfully install a local model but encounter an error when sending a message. The issue was related to inconsistencies between the chat layer and the model manager, particularly when resolving the selected model, checking installation status, finding the local model path, or initializing local inference.
Version 3.0.1 improves selected-model resolution, installed-model lookup, model path resolution, model loading, inference startup, chat routing, fallback behavior, and error reporting.
As a result, local chat on mobile devices is now significantly more reliable.
The model download experience has been improved to provide clearer feedback.
Users can now more easily understand what is happening while a model is being downloaded. The application can display download percentage, current progress, download speed, installation state, completion state, and failure state.
This is especially useful for larger AI models where downloads may take longer or require significant storage space.
The internal model download system has also been updated.
OpenMindAI can now more consistently distinguish between models that are not installed, preparing for download, downloading, verifying, installing, installed, loading, ready, or failed.
This improves the accuracy of the interface and reduces situations where the user sees an outdated or incorrect model status after a download has completed.
The model recommendation system has been expanded to make better use of device hardware information.
OpenMindAI can consider system RAM, CPU capability, GPU availability, VRAM, storage space, operating system, device type, and model size when determining which models are suitable for the current device.
This helps avoid recommending models that are too large or demanding for lower-resource hardware while allowing users with stronger devices to discover more capable options.
The setup and model-selection experience has been improved so OpenMindAI can show multiple recommended models rather than limiting users to a single suggestion.
Different models can be recommended for different purposes such as general chat, coding, reasoning, vision, lightweight local usage, and advanced workloads.
This gives users more flexibility when deciding which AI capabilities they want to install on their device.
The user-facing model system has been redesigned to make model selection easier.
Instead of requiring users to understand complicated provider names, original architecture names, quantization formats, and internal filenames, OpenMindAI focuses more on model purpose, supported capabilities, compatibility, size, and expected use.
OpenMindAI model families can be presented through simplified names such as OpenMindAI Nano, OpenMindAI Swift, OpenMindAI Core, OpenMindAI Titan, OpenMindAI Reasoning Mini, OpenMindAI Reasoning, and OpenMindAI Lens.
The underlying technical model information can still be used internally when required.
Model cards have been updated to provide clearer information.
The interface now focuses on the OpenMindAI model name, intended purpose, supported capabilities, recommended use, device compatibility, download size, installation status, and performance class.
Technical information that does not help most users make a decision has been reduced from the main interface. This results in a cleaner and easier-to-understand model manager.
The OpenMindAI v3 model architecture has been expanded to support additional compatible open-weight AI model families.
The underlying system can work with compatible models originating from ecosystems such as OpenAI open-weight models, Google Gemma, Meta Llama, Alibaba Qwen, and other supported open model families.
Where appropriate, OpenMindAI can expose these models through its own simplified model and capability system while preserving the required technical metadata internally.
Local and offline AI remain central to the OpenMindAI platform.
Once the required runtime and model files are installed, supported workloads can run directly on the user's device without requiring a continuous internet connection.
This allows users to run local chat, reasoning, coding, document analysis, and other supported workflows while keeping more of their processing on their own hardware.
Internet access may still be required for initial model downloads, software updates, web research, online integrations, connected APIs, and features that explicitly depend on external services.
The chat interface has been improved across both desktop and mobile versions.
The conversation experience is cleaner and more responsive, with improved model awareness, better loading states, clearer errors, improved input behavior, Markdown rendering, code blocks, copy actions, and better assistant response presentation.
The chat system now works more closely with the automatic model router, allowing tasks to be handled by suitable installed models while still supporting manual model selection when required.
Several internal changes were made to improve local chat reliability.
OpenMindAI now handles cases more consistently when a model has just finished downloading, has not yet been loaded, becomes unavailable, requires a different capability, needs inference to restart, or has an incomplete installation.
These improvements reduce failed requests and make local AI feel more stable during normal use.
OpenMindAI v3 expands support for reasoning-focused models and workflows.
Compatible models can be used for complex problem solving, programming, mathematics, scientific analysis, technical reasoning, research, structured planning, and multi-step tasks.
When an appropriate reasoning model is installed, OpenMindAI can route suitable requests to it automatically.
Coding support continues to improve in OpenMindAI.
The model capability system can identify models suitable for code generation, debugging, code explanation, refactoring, software architecture, documentation, repository analysis, and development workflows.
The v3 architecture also provides a stronger foundation for more advanced coding-agent features in future releases.
OpenMindAI v3 includes an expanded foundation for agent-based AI workflows.
The internal capability system can distinguish models that are suitable for tool usage, multi-step task execution, research agents, coding agents, document workflows, and automated reasoning.
This allows future OpenMindAI releases to provide more advanced agentic behavior without treating every model as a standard chat model.
Vision-capable models can now be better represented through the OpenMindAI model capability system.
Compatible models can support image understanding, screenshot analysis, UI analysis, diagram interpretation, visual document analysis, and image-based questions.
Availability depends on the installed model and the capabilities of the current device.
OpenMindAI includes a dedicated image-generation workflow.
Image generation is separated from ordinary text chat to provide a more focused experience for visual creation. The architecture supports image prompts, generated-image previews, workflow history, and dedicated image tools.
Availability depends on the installed or connected image-generation engine.
Document-related workflows have been expanded in OpenMindAI v3.
Users can work with longer and more structured content using dedicated document functionality rather than relying only on a standard chat conversation.
Supported workflows can include document writing, rewriting, summarization, analysis, research documents, technical documentation, professional reports, and long-form content generation.
OpenMindAI includes dedicated PDF-oriented workflows for creating structured documents.
This can be used for reports, research documents, technical documentation, academic content, professional documents, and other long-form material intended for PDF output.
Word-compatible DOCX generation is also part of the OpenMindAI document workflow.
This allows users to create editable documents for research, reports, academic work, professional communication, documentation, and general writing.
Markdown remains an important format within OpenMindAI, especially for developers and technical users.
It can be used for project documentation, GitHub content, release notes, research notes, technical writing, and developer-oriented workflows.
OpenMindAI v3 continues to expand research-oriented functionality.
The architecture supports document analysis, research assistance, technical investigation, long-form synthesis, structured research output, and reasoning-based research workflows.
When internet-based information is required, online research capabilities can complement local AI models.
OpenMindAI is also moving toward a more project-based workflow.
Projects can help organize conversations, documents, research, generated content, AI workflows, and related files in one place.
This makes it easier to maintain context for long-running work instead of treating every conversation as an isolated session.
Privacy remains one of the core principles of OpenMindAI.
When local models are used, prompts can be processed directly on the user's device, documents can remain local, conversations can be stored locally, and installed models remain under the user's control.
Local workflows do not need to send user content to third-party AI providers unless a feature specifically depends on an external service.
This separation between local and connected functionality gives users greater control over how their data and AI workloads are handled.
Version 3.0.1 includes multiple performance-related improvements across the application.
Model state synchronization, model selection, application startup, chat initialization, model loading, mobile navigation, user-interface state updates, and download handling have all been improved.
These changes are intended to make the application feel more responsive and reduce unnecessary delays during local AI workflows.
General stability has been improved across desktop and mobile.
Application startup, local inference initialization, chat execution, model manager synchronization, model downloads, model installation, mobile state handling, error recovery, and cross-platform behavior have all received fixes and improvements.
These changes make OpenMindAI v3.0.1 a more reliable release for daily use.
Version 3.0.1 resolves a number of issues discovered after the v3.0.0 release.
The mobile chat model error has been fixed so installed local models are less likely to fail when a user sends a message. Installed-model recognition has been improved to prevent downloaded models from incorrectly appearing as missing. Model-routing state synchronization has also been improved so manual selections and automatic routing work together more consistently.
The model loading flow has been updated to better handle cases where a model must be initialized before a conversation begins. Local model path resolution has been improved, reducing failures when the application tries to locate downloaded model files.
Download-status synchronization has also been fixed so the interface updates more reliably after a model finishes downloading or installing.
The recommended-model interface now handles multiple recommendations more consistently, while model capability information has been cleaned up to make each model easier to understand.
Mobile layout issues, chat loading states, routing fallback behavior, and model-manager state updates have also been improved.
OpenMindAI can now provide clearer errors when something goes wrong with a local model.
The system can better distinguish between a missing model, incomplete installation, loading failure, unsupported capability, hardware incompatibility, unavailable runtime, or inference failure.
This makes troubleshooting easier and reduces situations where users only receive a generic error message.
Several user-interface and usability improvements are included in this release.
Model selection is cleaner, capability descriptions are easier to understand, download information is clearer, model status indicators are more accurate, mobile navigation has been refined, and setup workflows have been simplified.
The goal is to make OpenMindAI easier to use for people who want local AI without requiring them to understand the technical details behind every model.
The desktop version continues to receive improvements around local model management, hardware detection, model routing, runtime handling, local chat, and model state synchronization.
The v3.0.1 architecture is designed to keep the desktop experience consistent across Windows, macOS, and Linux while allowing platform-specific runtimes and hardware acceleration to be handled where required.
The mobile application received some of the most important fixes in this release.
Local chat is more reliable, automatic model routing has been improved, model download progress is clearer, download speed can be presented, installed-model detection is more accurate, capability information is easier to understand, and hardware-aware recommendations are more useful.
The mobile interface also includes improvements to chat, image workflows, document workflows, and responsive layouts.
OpenMindAI continues to maintain a separation between local AI execution and optional connected services.
Version 3.0.1 improves internal handling of local model paths, model states, application storage, local inference, and online functionality.
Supported local AI workflows are designed to run without requiring external AI providers unless a selected feature specifically depends on an online service.
The v3 architecture continues to unify OpenMindAI across Windows, macOS, Linux, Android, and iOS.
Platform-specific implementations may still differ depending on hardware and runtime requirements, but the model capability system, routing architecture, model management experience, and overall workflow are being standardized across the ecosystem.
OpenMindAI v3.0.0 introduced the new generation of the platform, including the mobile application, expanded model architecture, hardware-aware recommendations, local AI capabilities, automatic model routing, document tools, image workflows, and cross-platform design.
OpenMindAI v3.0.1 strengthens that foundation with improved mobile chat reliability, better model routing, more accurate installed-model detection, improved model loading, better download progress information, multiple recommended models, cleaner model cards, stronger fallback behavior, more reliable local inference, mobile UI improvements, desktop stability improvements, improved error handling, and general performance fixes.
OpenMindAI v3.0.1 targets Windows, macOS, Linux, Android, and iOS.
Feature availability, model compatibility, and local AI performance may vary depending on the operating system, available RAM, CPU, GPU, VRAM, storage capacity, and device thermal limitations.
Larger models may provide stronger capabilities but also require significantly more hardware resources. Mobile devices will generally perform better with smaller optimized models unless the device has enough resources to support larger workloads.
Users running OpenMindAI v3.0.0 are strongly recommended to upgrade to v3.0.1.
The update is especially important for mobile users because it includes major improvements to chat execution, model routing, local model detection, model loading, download handling, recommendations, and overall application stability.
Desktop users will also benefit from improved model state synchronization, routing, local inference reliability, performance, and cross-platform fixes.
OpenMindAI v3 provides the foundation for continued development across local AI, multimodal workflows, agentic AI, coding assistants, research, voice capabilities, documents, model routing, hardware acceleration, and cross-device experiences.
Future releases will continue improving the balance between local privacy, offline functionality, performance, accessibility, and optional connected AI capabilities.
Thank you for using and supporting OpenMindAI.
OpenMindAI is being developed as a practical, privacy-focused AI ecosystem that gives users greater control over where and how their AI runs. The platform combines local and offline AI with desktop, mobile, multimodal, document, research, and connected workflows in one growing ecosystem.
OpenMindAI v3.0.1 represents another important step toward making powerful AI more accessible, private, and usable directly on the devices people already own.
Website: https://www.openmindai.org
Version: 3.0.1
Previous Version: 3.0.0
Release Series: 3.x
Platforms: Windows, macOS, Linux, Android, iOS
OpenMindAI v3 introduces the next major stage of the OpenMindAI ecosystem, expanding from a desktop-focused local AI application into a cross-platform AI platform for Windows, macOS, Linux, Android, and iOS.
Version 3.0.0 introduced the major new architecture, mobile applications, expanded local AI capabilities, model management, multimodal tools, document workflows, and automatic model routing.
Version 3.0.1 focuses on stability, model-routing reliability, mobile chat fixes, model download experience, and general usability improvements.
OpenMindAI now targets:
The goal of the v3 architecture is to provide the same OpenMindAI experience across desktop and mobile while allowing each platform to use hardware-appropriate local AI models.
The model system received a major redesign.
OpenMindAI now uses simplified OpenMindAI model names instead of exposing complicated provider, architecture, and quantization information throughout the normal user interface.
Model families include:
Different models are optimized for different hardware and workloads.
The v3 model architecture was expanded to support additional openly available model families and compatible local AI models.
The underlying ecosystem can use models originating from open model families including technologies from:
Provider-specific technical names are abstracted where appropriate so users can select models based primarily on capability instead of model implementation details.
Models can now expose the tasks they are designed to perform.
Capabilities may include:
This makes it easier to understand which model should be used before downloading or activating it.
Added an improved model-routing architecture.
OpenMindAI can select an appropriate installed model based on:
This reduces the need to manually change models for every task.
Manual model selection remains available when users want direct control.
OpenMindAI can use device capabilities when determining appropriate local models.
The model system is designed to consider factors such as:
This helps prevent unnecessarily large models from being recommended to devices that cannot run them efficiently.
The local Model Manager was expanded with:
Users can install additional models whenever they need more capabilities.
The setup workflow now provides recommended models instead of forcing users to manually determine which models should be installed.
Multiple compatible recommended models can be shown based on the device rather than displaying only a single recommendation.
A lightweight general-purpose model can be installed first so users can begin using local chat quickly and install specialized models later.
The chat system received a major interface and architecture update.
Improvements include:
OpenMindAI v3 expands reasoning-oriented workflows.
Supported model configurations can be used for:
Dedicated reasoning models can be selected automatically when appropriate.
Expanded support for software-development workflows.
OpenMindAI can assist with:
The architecture also prepares OpenMindAI for more advanced coding-agent capabilities.
Added and expanded multimodal AI workflows.
Supported models can analyze:
Vision-capable models are represented through the OpenMindAI capability system.
Added a dedicated image-generation experience.
The interface separates image-generation workflows from normal text chat, making it easier to create and manage generated visual content.
Expanded document workflows.
OpenMindAI can support document-oriented tasks including:
Dedicated document interfaces were added to make document generation different from ordinary chat.
Added dedicated document generation workflows for creating downloadable document formats including:
This provides a more structured environment for reports, research material, professional documents, and other long-form content.
Expanded OpenMindAI research capabilities.
The architecture supports workflows such as:
Online research functionality can complement local models when internet-based information is required.
The v3 architecture includes support for voice-oriented AI functionality.
Supported workflows can include:
Availability depends on the platform and installed capabilities.
OpenMindAI projects provide a way to organize related:
This creates a persistent workspace instead of treating every conversation as an isolated chat session.
OpenMindAI continues to follow a local-first architecture.
When using local models:
Internet access is still required for functionality such as initial model downloads, updates, web-based research, or explicitly connected online services.
Version 3.0.1 is a stability and usability update built on top of the major v3.0.0 architecture.
Fixed issues where a locally installed model could fail when a message was submitted from the mobile chat interface.
Improvements were made to:
Improved automatic model routing across OpenMindAI.
The routing system now handles installed models and supported capabilities more reliably.
Improvements include:
Improved the model download interface.
Model downloads can now expose clearer progress information including:
This makes large local model downloads easier to understand and monitor.
The model setup flow was improved so that compatible recommended models can be presented together rather than restricting setup to only one model.
Users can better understand which models are recommended for:
Model information was simplified for normal users.
Instead of emphasizing internal information such as:
the interface focuses primarily on:
Technical metadata can remain available internally where required.
Improved the mobile application interface across major AI workflows.
Updates include:
Improved detection of models that are already downloaded and installed.
This prevents the application from unnecessarily treating an installed model as unavailable or attempting to repeat setup steps.
Improved model installation state management.
The application now handles transitions between:
more consistently.
Multiple internal improvements were made across the desktop and mobile architecture.
These include:
OpenMindAI Desktop v3.0.1 includes packaged Windows x64 distribution support.
Desktop bootstrap and setup architecture also covers:
Platform-specific runtime and hardware capabilities may differ.
The local-first architecture continues to minimize unnecessary external transmission of user content.
v3.0.1 improves internal state handling around local models and keeps local AI functionality separated from optional connected or online functionality.
With v3.0.1, OpenMindAI is evolving into a unified AI environment covering desktop and mobile devices.
The ecosystem now combines:
Users running OpenMindAI v3.0.0 are recommended to upgrade to v3.0.1 for improved model routing, mobile chat reliability, model installation handling, download progress reporting, and general stability improvements.
OpenMindAI is being built as an open, practical, privacy-focused AI platform that allows users to run intelligent models directly on their own hardware while still providing access to more advanced connected capabilities when needed.
More improvements are already being developed for future OpenMindAI releases.
Website: https://www.openmindai.org/
Release: OpenMindAI v3.0.1
Previous Major Release: OpenMindAI v3.0.0
Platforms: Windows, macOS, Linux, Android, iOS
Every stored Windows, macOS, Linux and Android package is listed independently with its real filename, architecture/ABI, size, source and install command.
.\OpenMindAI_3.0.1_x64-setup.exesha256:321630769183f096115f1149a249ce8b2f5919c8203102f6af2ec6fc2e4b71f4.\OpenMindAI_3.0.1_x86-setup.exesha256:4f724b09d41c9c4dae6e9d038e85f7c960a3cf7ba2fb619a6be17367892ce8ddmsiexec /i "OpenMindAI_3.0.1_x64_en-US.msi"sha256:79f50a742e534cedb799e4e508e3a6d03f22f27087bd7989bb281658c6ca5beemsiexec /i "OpenMindAI_3.0.1_x86_en-US.msi"sha256:dddd20811e0e3ccb69f71dc4a5db13184ea4954bc6cbcb49d3cee331c94a98e5.\OpenMindAI-Setup.batsha256:16f7e7cfb346778b708ee16d0f40d2ac8a3724eddecb4e5ab3e40c503a6b7b39chmod +x "OpenMindAI-Setup.command" · "./OpenMindAI-Setup.command"sha256:8b0c85a28ce63b958a2a5057c28e1fd001b5e457dbcd5c047263cd9e01712c82open "OpenMindAI_3.0.1_aarch64.dmg"sha256:d098edfbdb0aa2b8e9c946a544cb025c744734614642994b98f93a5da35992c4open "OpenMindAI_3.0.1_universal.dmg"sha256:2557fa8899152f1ce964d9c1dc3af01496ea3c1c69130224f626ce83f4bab5f6open "OpenMindAI_3.0.1_x64.dmg"sha256:fb69088ab260d4f0be9f1c88deade384384ae89ba99c4915159f37e1e428efbbchmod +x "openmindai-setup.sh" · "./openmindai-setup.sh"sha256:5c700954d0d253ffd1a9e59ed0a717c210dcc1921134f6ae97fd2dc69a3e9e0fchmod +x "OpenMindAI_3.0.1_amd64.AppImage"
"./OpenMindAI_3.0.1_amd64.AppImage"sha256:9d87963e6b28f2275453190ccd2f20f24535c3eb7cd61eebe79eb76a6f0400bfchmod +x "OpenMindAI_3.0.1_i386.AppImage"
"./OpenMindAI_3.0.1_i386.AppImage"sha256:c49347518bd9f854cddbdabb5c35095c6b7ac963500e50c9cf79cbc3dfb48cf7sudo apt install "./OpenMindAI_3.0.1_amd64.deb"sha256:486b9043530e76fa9e5f4bfbd75b3c64b7e38d7b4e54772f95bedbb7b3ad9313sudo apt install "./OpenMindAI_3.0.1_i386.deb"sha256:298cff6c59f5c95267eea847f83f689247c67f40429a7a1ec196fad49c294711adb install "app-arm64-v8a-release.apk"sha256:2016e230fb76b16432b2e151b95dee2926bdd56451b2608da6c64fc682107232adb install "app-armeabi-v7a-release.apk"sha256:224c218ad48c094c573138f713a51763d70a55400ac93c9020549b0a7b199b1eadb install "app-x86_64-release.apk"sha256:c3f19f0dcf490a08daa848ef6b3cd17628bc0df89f19093d2eee199601a7110csha256:fedd5c380d6070d9b0b381d0446d9ddfac4e20b006e8f507de196a06ee6d4bb7Compare these values with the downloaded packages before installation.
app-arm64-v8a-release.apk: sha256:2016e230fb76b16432b2e151b95dee2926bdd56451b2608da6c64fc682107232 app-armeabi-v7a-release.apk: sha256:224c218ad48c094c573138f713a51763d70a55400ac93c9020549b0a7b199b1e app-x86_64-release.apk: sha256:c3f19f0dcf490a08daa848ef6b3cd17628bc0df89f19093d2eee199601a7110c MacOS-SHA256SUMS.txt: sha256:6e4507a2d75c92b7a944a722dd95432a6bfa99efe71e03bde48dc87f971b7fc5 openmindai-mobile.ipa: sha256:fedd5c380d6070d9b0b381d0446d9ddfac4e20b006e8f507de196a06ee6d4bb7 OpenMindAI_3.0.1_aarch64.dmg: sha256:d098edfbdb0aa2b8e9c946a544cb025c744734614642994b98f93a5da35992c4 OpenMindAI_3.0.1_amd64.AppImage: sha256:9d87963e6b28f2275453190ccd2f20f24535c3eb7cd61eebe79eb76a6f0400bf OpenMindAI_3.0.1_amd64.deb: sha256:486b9043530e76fa9e5f4bfbd75b3c64b7e38d7b4e54772f95bedbb7b3ad9313 OpenMindAI_3.0.1_i386.AppImage: sha256:c49347518bd9f854cddbdabb5c35095c6b7ac963500e50c9cf79cbc3dfb48cf7 OpenMindAI_3.0.1_i386.deb: sha256:298cff6c59f5c95267eea847f83f689247c67f40429a7a1ec196fad49c294711 OpenMindAI_3.0.1_universal.dmg: sha256:2557fa8899152f1ce964d9c1dc3af01496ea3c1c69130224f626ce83f4bab5f6 OpenMindAI_3.0.1_x64-setup.exe: sha256:321630769183f096115f1149a249ce8b2f5919c8203102f6af2ec6fc2e4b71f4 OpenMindAI_3.0.1_x64.dmg: sha256:fb69088ab260d4f0be9f1c88deade384384ae89ba99c4915159f37e1e428efbb OpenMindAI_3.0.1_x64_en-US.msi: sha256:79f50a742e534cedb799e4e508e3a6d03f22f27087bd7989bb281658c6ca5bee OpenMindAI_3.0.1_x86-setup.exe: sha256:4f724b09d41c9c4dae6e9d038e85f7c960a3cf7ba2fb619a6be17367892ce8dd OpenMindAI_3.0.1_x86_en-US.msi: sha256:dddd20811e0e3ccb69f71dc4a5db13184ea4954bc6cbcb49d3cee331c94a98e5 output-metadata.json: sha256:0ec9c8ee9667fd3155091546cf3c6001ed5042bc9c4c95055099309dbdaa4355 release-output-SHA256SUMS.txt: sha256:6fd5ca235a7c6ebe3d9f95a8d3fc0b261b7f3712a1faebe65bab4b893f6a05a9 Runner.app.zip: sha256:9a45a7722f2d7e14459a487fe640e29fd2017dc2a1cf33c6ee3a24bc27817b43 Windows-SHA256SUMS.txt: sha256:bf27f37025b7b466aca3540bd891fc4e43006337897212d060a91c79993fc51b