What is Worldku?
One memory, running every app
Worldku is a personal memory runtime for AI apps: it organizes your chats, journals, voice, games and work records into one structured memory you own. Apps no longer keep isolated memories — with your authorization they read task-relevant context, and the same memory serves every AI app you trust.
Memory always belongs to the user: review, disable, export or delete it at any time.
Memory layer
Taking memory out of the app
Most AI apps treat memory as a local feature: one app remembers one version of you, the next starts from zero. Worldku makes memory an independent layer — structured context of facts, experiences, timelines and source references, owned by the user; apps read by authorization and write back by permission.
01
Memory belongs to the user
Worldku is the single place to review, export and delete memory. Apps may use memory within authorization, but never become its owner.
02
Records become structure
The system doesn't merely store raw text — it distills facts, experiences, preferences, relationships and long-term patterns, a form built for long-term use.
03
Read scope is defined by authorization
An app receives only the memory scope it was granted. Memory is governable context, not invisible material hidden inside a prompt.
Three layers
From source records to runtime context
Worldku organizes user data into three layers, each answering one question: how raw records are kept, how memory is distilled, how apps consume it.
01
Source records
The raw material of chats, journals, voice, games and app events. This layer belongs to the user first — review, export, delete, or exclude it from long-term memory.
02
Structured memory
The long-term form distilled from raw records: not transcripts, but facts, experiences, timelines, preferences and relationships — recurring goals, changes of state over time.
03
Runtime context
When an app requests memory, the runtime selects only a small, task-relevant slice within the authorized scope — memory accumulates for years without becoming a heavy dump.
Across apps
One memory, many applications
Worldku is not another chat window — it is the shared memory layer behind different AI apps: a writing app understands your style, a music app your taste, a game your strategy, travel your places, learning your progress, and work tools your long-running state. Apps focus on their own experience; Worldku runs memory, authorization, accounts and billing underneath.
The Developer Preview will open gradually — letting AI clients like Claude, Codex and Cursor plug into the same memory with the user's authorization.
Runtime
One Effort interface, a multi-stage orchestration system
Most apps hand an input straight to one model; the Worldku Runtime first understands the task, recalls relevant memory and analyzes behavioral state, then renders the result. For users and developers the interface is always a single action — choose an Effort, and Worldku decides how much reasoning, memory and orchestration it takes.
01
Memory Recall
The runtime first retrieves context from the user's authorized memory; apps only declare the memory scope a task needs.
02
Behavior Analysis
At higher Effort, it first reads the behavioral signals behind an input — a quick question or a deep review — then sets the depth, pace and structure of the output.
03
Multi-Agent Orchestration
Complex requests are split across internal agents and model paths: understand, retrieve, analyze, generate. The user still sees one complete, natural result.
04
Final Rendering
Merges the task goal, app rules, user memory, behavior analysis and model output into the final expression.
Effort is not a model name — it is the runtime's operating depth. Low Effort is faster and lighter; high Effort enables more memory, deeper analysis and more complex internal orchestration.
Pricing
Priced by operating depth
Worldku pricing is determined by three things: available Effort depth, monthly usage allowance, and the completeness of memory capabilities. Start accumulating memory for free; subscriptions unlock a deeper runtime.
Free
$0/mo
Start accumulating memory
- Basic memory management
- Base Effort
- Limited daily allowance
- Data review, export and deletion
Pro
$8/mo
or $80/yr
Everyday use across apps
- More monthly credits
- Higher Effort
- Cross-app memory reads
- Voice usage allowance
- Full memory management
Max
$20/mo
or $200/yr
Heavier runtime workloads
- Highest credit allowance
- Highest available Effort
- Deeper memory recall
- Multi-stage runtime orchestration
- For reviews, analysis, creative work and long-term records
Payments are securely processed by Creem. Exact prices, allowances and available features are shown at checkout. Subscriptions can be managed or canceled anytime from your account.
FAQ
Frequently asked questions
Q1Is Worldku an AI chat app?
No. Worldku is the memory and runtime layer behind AI applications. The main site is where you manage memory, account, voice, usage, billing, export and deletion.
Q2What does Worldku store?
Records the user owns, and the structured memory organized from them. Records can include chats, journals, voice activity, gameplay, work status and app behavior; structured memory includes facts, experiences, timelines and source references.
Q3Can apps read all of my memory?
No. Apps use relevant memory only within the user's authorization and product scope — every read is bounded by the granted scope.
Q4What is Effort?
Effort is the runtime's operating depth. It decides how much reasoning, memory, behavior analysis and internal orchestration a request uses. Users never need to care which underlying models were involved.
Q5Can I delete or export my data?
Yes. Memory belongs to the user — review, export, disable or delete it in Worldku at any time.
Q6How do I manage or cancel a subscription?
Subscriptions are processed by Creem and can be managed, upgraded or canceled anytime from the billing page in your account; after cancellation, benefits remain until the end of the current billing period.
Q7Will developers be able to plug in?
Yes. The Developer Preview will open gradually. The goal is to let apps use Worldku's memory and runtime capabilities without building accounts, memory, billing, voice and data governance from scratch.