AI Twin Brain Simulator

Every account's AI Twin has a unique brain: a living neural map built from what your twin has learned while working with you. The Brain Simulator lets you open that brain, watch it light up while your twin produces, replay how it thinks, browse its memories like a wiki, and — if you want — show a reduced version of it on your public profile.

Nothing in the simulator changes your twin. It is a window, not an editor.

How your brain maps to your work

The brain uses three familiar anatomical ideas, each backed by your real Chainabit data:

Brain concept	What it is in Chainabit
Neurons	The concepts your twin has learned — your bits and topics, plus the tools you use, your traits, people you work with, and memories
Synapses	The connections between concepts — your chains and the learned links between ideas that fire together. Stronger, more reliable connections render thicker and brighter
Regions	Your chainies — the areas of your life and work that group related neurons and synapses, each with its own color and progress

And one more: activity. Your brain shines while your twin produces — recent runs, AI sessions, and signed chain periods light up the matching neurons and synapses in real time, fading as they age out of the activity window (from 5 minutes up to 24 hours).

Simulate thinking and trace decision paths

Two interactive features run in the app, deterministically, and for free — they replay your graph, they do not call an AI model and they never spend credits:

• Simulate thinking — pick any neuron as a seed and watch an activation wave spread through its synapses, strongest connections first. Run it twice, get the same wave: it is a faithful replay of your brain's current wiring, not a generated story.
• Decision paths — select two or more concepts and trace the path your brain connects them through, step by step, with the strength of each hop.

When a wave or a path surprises you, that is the point: it shows which ideas your twin considers neighbours, and how strongly.

The memory graph — your twin's wiki

Your twin's memories form a second map, browsable like a wiki:

• Each memory is a page: its type, content, importance, relevance, and tags.
• Consolidation links show when several smaller memories merged into a stronger one — follow the link to see what the summary was built from.
• Tag clusters group related memories so you can scan themes at a glance.
• Each memory's detail page shows provenance (where it came from) and recent behavioural evidence, so a claim is never a black box.

You stay in control: any memory can be flagged as wrong or permanently forgotten via the Twins API.

AI narration

For any focus concept, decision path, or activation wave, you can ask for a short AI-written narration — a second-person story of what that region of your brain means, in your language (English, Turkish, Spanish, French, Arabic, German, or Portuguese).

Narration is the only part of the Brain Simulator that uses an AI model, and it is billed transparently:

• Each narration is charged to your wallet, based on the model tokens it used. The response tells you exactly what was charged.
• If your wallet cannot cover it, the request fails with 402 insufficient_credit and nothing is charged — top up via the Wallet API.
• Each plan has a daily narration budget. When it is exhausted you get 429 and the budget resets the next day.

Plan	Narrations per day
Bitter	5
Chainer	25
Architect	100

Narrations are grounded in your graph only — node labels and their connections. Your raw memories are never sent to the model.

Plan availability

Capability	Free	Bitter	Chainer	Architect
Brain graph, simulate thinking, decision paths	One-time preview	✓	✓	✓
Live activity ("your brain shines")	—	✓	✓	✓
Productivity overlay (regions & chain synapses)	—	✓	✓	✓
Memory graph & memory wiki	—	✓	✓	✓
Chain impact	—	✓	✓	✓
AI narration	—	5/day	25/day	100/day

The Free preview is real and one-time. Free accounts can open their actual brain once — real data, up to 80 nodes — to see what their twin has learned. After that single look, the simulator locks until you upgrade (a second attempt returns brain_sample_already_used). The app can check availability beforehand, so you will not waste the preview by accident.

Public brain showcase

You can optionally showcase a reduced version of your brain on your public profile at GET /public/twins/{username}/brain — a conversation piece, not a data dump:

• Opt-in and off by default. Turn it on with the publicBrainShowcase twin preference (Twins API); turn it off the same way at any time.
• Reduced data only. Visitors see your top topics, tools, and traits with their connections — capped in size, with no memories and no people, ever.
• Your profile must be public. If your profile is private, your brain is never shown, even with the toggle on.
• Changes — including switching the showcase off — can take a few minutes to propagate.

Privacy guarantees

• Your brain is yours. Brain, memory, and activity endpoints only ever show you twins in your own scope; other people's personal twins are invisible to you.
• The showcase shares the minimum. Top topics, tools, and traits only — never memories, never people, and only while both your profile is public and the toggle is on.
• Narration never sees raw memories. It is built from concept labels and connection strengths in your graph, nothing more.
• No surprise AI spend. Everything except narration is free and deterministic. Narration is always charged to your own wallet with a clear per-plan daily cap — there is no background AI activity in the simulator.
• Memories are correctable. Flag a wrong memory to demote it, or forget it permanently — forgotten items are not re-derived.

API reference

Endpoint-level details — paths, parameters, request and response shapes, and error codes — are in the Twin Brain API reference.