In this page, we will provide continuous updates for our software, hardware designs, 3D printable CAD files, download links, any change in payment details or any important updates during development. Especially, in case someone makes payment without any contact details and we don’t receive the payment or any payment related issues, we will give update on this page.

Last Update: 29th August 2026:(For Active Subscribers)

Our software has reached the final stages of development. As usual, the temporary download link will be available on download tab for download in case if are an active subscriber and if you don’t have access to our Github private repository.

Update: 8th July 2026:(IMPORTANT UPDATE)

The head assembly of the robot has resumed back. We already assembled it partly and due to various reasons like polishing parts and delivery delays, it was paused temporarily. Today we are resuming back the head assembly and it can still take a few weeks or more. Once the head assembly is complete and all servos are tested and electronic gadgets are verified to be working, the silicon work will start for Silicon face mask. Note that any robot related CAD files will be available publicly or in private repos ONLY after we complete the full robot along with robust testing. Our private repos will contain primarily the updated versions of the Desktop app and Android app if we apply any changes. However, from time to time we will upload CAD files and design files in temporary download links as usual so that the “ACTIVE SUBSCRIBERS” can download the files. Also, note that whenever we receive payment for the monthly subscription fees, usually we update any latest updates in our temporary download link under the “Download” tab around 1 to 2 weeks of time as per our old standards how we have been uploading the development updates so far. Github is primarily used for maintaining a consistent record of verified and well tested documents, codes and CAD files where as the continuous updates will be usually available in temporary download links as usual.

Update: 6th July 2026:

We have now most of the code uploaded in the private repos. So the active subscribers can get access if they want. Uploading full code to public repository can still take a couple of months and we will decide that after careful verification of all associated software components. Now, we have 2 parallel running back end and front end apps. Both apps have similar architecture. The original app “AGI IN ACTION” which is partially uploaded to Github is heavily dependent on RAG consisting of both TEXT RAG and VISION RAG. It has one desktop app and one Android app which are connected via TailScale. The second project is entirely designed to depend on continuous LORA training which is still buggy and not reliable. Due to complexity and less reliability we are now focussed entirely on RAG based approach and planning to use RAG even in humanoid robot at least in the first version or V1. The first version of the robot will be complete version of the robot with just the head servos and fully body silicon layer. It means all other robot joints will use heim joints instead of any servos. It is enough to use to just RAG for V1. In version 2 or V2, we will add hand servos as planned earlier and then, in V3 we will add leg servos. Still it is roughly a 1 to 2 years of project. But we believe the first version V1 will be completed in next few months and we can try to finish the design and assembly by end of this year. If the full design is complete, then all polished CAD files will be available in private repos initially as usual.

Update: 29th June 2026:

The legacy version is available for download under “GitHub” tab. We have uploaded the legacy main file where FRP server client is used. We have now entirely switched from FRP Client server application to purely using Tailscale end to end encryption in the latest version. It is available in our private repo. So the active subscribers who joined recently or actively funding to the project must submit the Github username or email ID associated with Github through the “SUBSCRIPTION FORM” of our website as per their convenience so that we can send a READ ONLY invitation. The codes will not be available publicly immediately due to security reasons. It may take few months or even years to fully open source the latest version. But active subscribers can have READ ONLY access to the ongoing repository. Also, unlike earlier approaches of providing external download links, now all code will be centralised and contributed through Github based on our website “Terms Of Service”. All subscribers must strictly abide by our existing “Terms Of Service” and any future changes to stay compliant with current and future AI regulations.

Update: 7th June 2026

Status: Desktop app and Android app are mostly complete and functional. But full code is not updated to GitHub. This project is heavily memory dependent using multilayer memories, FRP Client server files along with many other open source projects interlinked with our project. So we want to make sure we don’t leak personal memories, IP addresses to public repos or don’t violate any community guidelines before uploading all codes online. It uses TEXT RAG and VISION RAG for lifetime memory along with rolling window context memory. Also, it is integrated with OpenClaw which has it’s own separate memory with the same dedicated Android app centralised one GUI based Desktop app built in pySimpleGUI. So there is no need to connect with external apps like Telegram, whatspp etc and you can use your own dedicated Android app privately and locally.

We will add continuous LORA training in future versions when humanoid robot will be somewhat ready and FULLY ALIVE. Now, it is using direct LM Studio API Endpoint locally. Humanoid robot is under development. We are using existing humanoid robot open source projects components whatever is available and adding our own polished components during fine tuning. For now, robot head assembly is in progress.

We are developing in phases or versions and so first version will have full robot in sitting posture in V1. In version 2 or V2 we will add hand servos just for hand movement and working on a computer by sitting on chair. In V3, we will add leg servos. In final V4 version we will add PD controllers and may add RL policies if absolutely necessary as last resort. Our goal is to primarily use PD and PID along with LLM generated commands via a PLANNER and EXECUTOR approach which is standard approach in robotics.