Start here

What Socria is

A thinking environment, not an answer machine — and how these docs are organized.

The idea#

Socria is built on a single conviction: the thinking should stay yours. Most AI products optimize for producing your answer. Socria optimizes for the quality of your own reasoning on the way to it — it asks before it answers, it notices the assumptions in your wording, and when you are working a math problem to learn it, it will guide you to the result rather than hand it over.

That stance is called Human-First AI, and it is not a tone of voice — it is enforced in the product. In Core 3.1 and Logos, questions are earned, never reflexive; conclusions stay yours everywhere. The map draws your claims, not its verdicts. And even Draft Space — the one surface built for prose — starts blank and stays yours: Socria offers rewordings of passages you wrote, and they land on the page only if you apply them.

The shape of the product#

Everything lives at one address — socria.app/chat — behind one model switcher. This is what the fullest of them looks like: the conversation on the left, a map of the thinking it produced on the right, kept in step with each other.

A reasoning environmentBalanced

I got an offer with a lot more money. I think I should take it — honestly I've stopped growing where I am.

Socria

Three years of the same problems is a real reason to leave. The money is a separate one — and you've put them in the same sentence.

…I guess I'm treating the raise as proof it's a step up.

Socria

That is the assumption worth testing. What would the new role teach you?

What are you thinking through?

Thinking Mapdeciding7 nodes

GraphSee how everything connects.

questionclaimassumptiontensionvalueevidence

The conversation and its map, side by side — as it actually runs

Four models share that address:

Core 3.1 — the conversation that remembers: thread memory, periodic syntheses of what you have worked out, and a running model of how you think. It opens with no account, for a first session kept in your browser.
Core 4 — the model that decides, each turn, which part of the work is yours and does everything else itself. It keeps a record of the reasoning — claims, objections, decisions, and whose each one is — so a long piece of work stops circling.
Logos 2 — the reasoning environment. The conversation runs beside a live Thinking Map that draws your claims, assumptions, tensions and evidence as you talk — and when what you describe is a system, the engine builds a model of it you can move, read, edit and undo: a surface, a mechanism, a set of equations, a fitted relationship, or a simulation of a black hole. With mathematics support, a Draft Space, and research grounded in real sources.
Logos 3.5 — Logos 2 with a workspace around it: charts, plans, comparisons and decisions made from your own words (what is new), and room for two people at once — one map, one set of models, and Socria as the layer between you, naming the connections and the disagreements without taking a side.

Core 2 retired on 2 October, and the original Logos has given way to Logos 2.

Core 3.1 and Core 4 are chat models: different depths of attention, and in Core 4’s case a different question — not how deep to go, but whose work this is. Logos 2 and Logos 3 are a different kind of thing — an environment — and it is the centerpiece of Socria One, the $15/month plan. Its free tier is not a trial of Logos 2 but Logos 2 itself, counted — 2 lines of thinking a month and 10 messages a day: what Socria One sells is how much a day and a month can hold, and what is carried between conversations, never a better version of any one of them.

What never changes#

Authorship. Your conclusions are yours. Socria asks, reflects, challenges and connects; it does not decide for you, and it does not write for you — even in Draft Space, where its Refine suggestions are proposals that land only if you apply them yourself.
The Answer Guard. While you are learning mathematics — working a problem to understand it — Socria will not hand you the result: not in the chat, not on the map, not on the Board. The door out (revealing the solution) is always yours to open, and it is never priced.
Transparency. You can always see where a thought on the map came from — the Trace move quotes you verbatim — and always tell Socria it read you wrong. Neither is ever gated.
Nothing held hostage. Reaching a free limit — the day’s chats or messages, or the month’s lines of thinking — pauses only what is new; it never hides, deletes, or locks what you have already thought. Your existing maps stay visible and interactive, the conversations you have begun stay open, and the day’s allowance comes back tomorrow.

Where to go#

What to use it forDeciding, learning, writing, working through math — matched to the right model and settings.The modelsCore 3.1, Core 4, Logos 2 and Logos 3.5 side by side — what each is for and how to switch.Socria Core 2Calm, restrained Socratic questioning. Retired on 2 October; Core 3.1 took its place.Socria Core 3.1The conversation that remembers, synthesizes, and follows your thinking across sessions.Socria Core 4Decides, each turn, which part of the work is yours — and contributes everything else.Socria LogosThe original reasoning environment — the conversation and a live map side by side. Superseded by Logos 2.Socria Logos 2The conversation, a live map, and models the engine builds from what you say — what you can build, what to ask for, and what you can do with one.Socria Logos 3The current Logos, now 3.5: the conversation, the map, models the engine solves and designs drawn in 3D from your words, in one workspace — and room for a second person.What’s new in Logos 3.5Charts, plans, calendars, comparisons, diagrams, number lines and decisions made from your own words beside the map — plus the experiences, a memory that tracks how your thinking changes, and what it will not do.Engineering in Logos 3Engines, mechanisms, circuits, structures, flows, flight, reactions and control — what to ask for, what the engine builds, and the numbers it computes, each checked — and CAD-style designs drawn in 3D.Dynamics and chaos in Logos 3Maps and flows: the road to chaos and Feigenbaum’s δ, a strange attractor, the Lorenz butterfly cut by a Poincaré section, and predator–prey cycles — each built live, each number computed.The Thinking MapNode types, relationships, the four lenses, and the four moves a node opens.MathematicsMath-aware maps, the Board, an interactive graph you can drive, and the Answer Guard.EconomicsSupply and demand, opportunity cost, the production frontier and AD–AS — drawn so you can move them.Depth, Personality & instructionsHow far the thinking goes, the one personality it talks in — the same in Core 4 and Logos — your own instructions, and what always stays fixed.Draft Space & groundingThe one surface built for prose — yours to write — and how nodes get grounded in real material.Socria OneThe free tier, the subscription, when Socria asks, access codes, and how billing behaves.Accounts & your dataSigned out vs signed in, what is stored where, sync, and deletion.