Private AI. Local power. Human control.
A place for your work.
AI on your terms.
Discover a local-first AI platform built around your projects, your knowledge, and the infrastructure you control.
a place to grow.
Inside FridayLocalAI
Start with what you want to do.
Organize your work
Friday’s Workshop brings projects, folders, and conversations into a shared workspace. Give ongoing work a clear home and return to it with the right context.
Learn more →Work with knowledge
Bring source material and scoped memory into your workflow. Understand the difference between storing a document and making its contents available to a task.
Learn more →Stay in control
Agents, skills, and tools operate within configured boundaries. Human authority, permissions, and evidence remain central to the work.
Learn more →Learn with Friday
Ask for help with the next step.
Use FridayLocalAI itself as a guide: describe what you want to do, ask for one step at a time, and check the result before continuing.
“I am new to FridayLocalAI. Help me start a practice conversation, explain the controls I can see, and wait for me after each step.”
Open the Help & How-To Guide · Request early access
The FridayLocalAI Store
A new chapter is taking shape.
FridayLocalAI is available through early-access inquiries. Contact the project to discuss access. There is no public installer, and the FridayLocalAI Store is not accepting purchases.
A local AI workspace for work that continues
Many useful AI tasks take more than one exchange. A research question becomes a collection of notes. A draft becomes several revisions. A technical problem needs a record of what was tried and what changed. FridayLocalAI brings that continuing work into a workspace with projects, folders, conversations, contextual guidance, and model controls. The purpose is to help you return to a task with its background intact and its next step clear. Start with something ordinary: organize meeting notes, compare two plans, outline an article, or ask questions about a document you are allowed to use. Keep the request small enough that you can evaluate the answer. As your work grows, the same habits apply: choose the right context, explain the outcome you need, and review what the application actually produced.
What local-first AI means here
Local AI concerns where computation happens and how the environment is operated. Private AI also raises questions about access, storage, connected services, and the people responsible for those decisions. FridayLocalAI treats those questions as part of the workflow. A local model can be one component of a larger application; it is not, by itself, a complete privacy or operating policy. Your deployment determines which models and services are available. Some workflows may use local resources while others depend on a configured external provider. Before using proprietary data, confirm the intended processing path with your operator. The website explains these choices without promising that every installation is disconnected, every feature is enabled, or every machine can run the same workload.
Begin with a guided practice session
An early-access participant can learn the interface by asking Friday for help inside a conversation. A useful first request is a guided tour focused on one task. Tell Friday what you can see, ask it to explain the next control, and report the result before continuing. Information icons provide another way to look up section-specific guidance when available. For example, type three fictional meeting notes and ask for a short action list. Check that the response preserves the facts and does not invent owners or deadlines. Ask for one revision, then return to the conversation to verify its history. This simple exercise introduces prompting, review, revision, and continuity without requiring private files, administrative changes, or a complicated integration.
Build context deliberately
Useful context comes in several forms. The conversation contains the current exchange. A project groups related work. An attachment supplies source material when the application can make its contents available. Memory Notes record guidance in a selected scope. These functions complement one another, but they are not interchangeable and should not be treated as automatic access to everything you have ever stored. When a question depends on a source, identify that source and ask for a checkable passage reference. When a preference should apply to future work, save it through the appropriate memory workflow and verify the result. When information is only relevant to the current question, keeping it in the conversation may be enough. The guides explain how to make these decisions without turning every detail into permanent guidance.
Match capabilities to the task
An Agent defines a role and associated configuration. A model generates a response. A Skill provides reusable guidance, while a Tool exposes an operation the runtime can allow or deny. Selecting a helpful name in the interface does not automatically make every capability available. Check the controls, permissions, and result for the work you actually intend to do. Text conversation, source analysis, image understanding, and media generation can require different resources. The requested Model selector and the active model attribution help you understand the response path. Media Studio provides a separate place for configured creative workflows. If the necessary option is missing, ask your operator about availability rather than assuming a prompt can enable it.
Choose a useful next step
If you are exploring the project, read the platform overview and compare local and connected workflows. If you already have access, begin with the Help guide and Quick Start. If you are planning an installation, contact the project with your intended tasks and environment. Early access is arranged by inquiry; the public website is not a self-service installer or an application account provider. The learning library offers background reading through clearly disclosed Amazon Associates links. Those third-party products are separate from the FridayLocalAI Store, which is not accepting purchases. You can read the web guides and download consumer documentation without buying a book. Choose the resource that answers your immediate question, then use a small practice task to turn the explanation into something you understand.