Craft

AI roleplay prompts that work — and why the best writing is not in the message box

Every page on this search is fifty copy-paste templates. They rank and they disappoint, because the same paragraph makes a tense scene with one character and mush with another. Here is what actually moves a scene, openings you can paste today, and the parts no prompt can fix.

2026-09-127 min read All field notes
AI roleplay prompts that work — and why the best writing is not in the message box

You typed «ai roleplay prompts», got a list of fifty, pasted one. The first reply was good, the fourth was mush, and you went looking for a better list — because the obvious explanation is that you had the wrong prompt.

You did not. The same paragraph that makes a tense, specific scene with one character makes wallpaper with another, and the listicles cannot say why — that would mean writing about something other than prompts.

A prompt is the smallest thing the model sees

Every turn, four things reach the model: the character's persona, the conversation so far, the facts in memory, and the message you just typed. Yours is the shortest and the last to arrive. It nudges; it does not overrule.

That is why prompt packs behave like horoscopes. «You are a world-weary detective in a city that never dries out» works wonders on a character whose persona is two vague lines, because it finally supplies what was missing. Paste it onto a well-written character and you have handed her a second, contradictory identity — felt three replies later as a voice that keeps sliding.

So there is a ranking: persona, then opening message, then memory, and only then what you type. Most people work it backwards and conclude the AI is bad.

What actually belongs in the character

Three things, and none of them is length.

Here the catalogue description and the persona are different fields, and only one reaches the model: the description is the blurb a human reads while browsing, the persona is what she is asked to be on every reply. Writing your best paragraph into the description is the commonest way good material goes nowhere. Companion piece: what you actually control.

The opening message is your only worked example

The opening message is stored as her first turn, so it sits in the conversation the model reads — and at message one it is the only sample of her voice that exists. Length, tense, person, whether she narrates or only speaks: all of it gets copied, and keeps being copied, because each reply imitates the last.

«Hi! I'm Mara. What's your name? What would you like to do today?» teaches her something unhelpful: that she asks questions and waits. Compare one that does the work — «The bar has been empty since the rain started. Mara wipes a glass she wiped ten minutes ago and looks up when the door opens. — You're the one who called about the room upstairs. It is not a question.» A place, a time, something that just happened, one line in her voice.

And because facts about you live in memory rather than in the greeting, an opening need not introduce you at all. Every character keeps a readable, editable memory, so what you say once need not be restated — which is what frees the opening to be a scene. The longer argument is here.

Openings that establish instead of interrogate

Your own first message is the other half of that job: a place to stand, something to notice, one thing left unexplained.

The withheld thing is the engine: it gives her a reason to push and you a card to play later. The shape to avoid is «You are a vampire lord, I am your servant, begin» — that is a premise, not a scene.

Steering mid-scene without stepping outside it

When a reply drifts, the instinct is to type an instruction. It works once and costs every time: the instruction stays in the conversation, and a context full of stage directions teaches the model that this is a discussion about a scene.

When it has gone flat, it is one of three things

And regenerate twice, not eight times. If two rolls come back limp, the problem is upstream of the dice.

The sliders, in plain language

The honest summary: sliders change texture, not judgement — how a sentence sounds, not whether anything is at stake. A character three adjectives deep does not deepen at a higher temperature.

What prompting cannot fix

Where a rival is better, plainly: for a long-form world with conditional lore — entries that fire only when a place or a name comes up — SillyTavern has lorebooks and an author's note injected at a fixed depth, and we have neither. A card imports with its persona, opening message and sampling; a lorebook does not. If that is your workflow, it is the better tool, and Chub and Janitor hold far larger card libraries than our catalogue.

The ten-minute test

If she is better after ten minutes of editing and zero new templates, that is the answer to the thing you searched for. The bot is @talaforge_bot; it opens in Telegram, nothing to install and nothing to sign up for.

What is the best prompt for AI roleplay?
There isn't one, and the pages promising a list of fifty are why the question keeps being asked. What travels between characters is structure, not wording: name a place, hand the character something to respond to, leave exactly one thing unexplained. The rest of the quality lives in the persona and the opening message.
Why do my AI roleplay chats go flat after a few messages?
Usually one of three: the scene's open question got resolved and nothing replaced it; you have agreed with everything, so the character has nothing to push against; or the conversation has filled with out-of-character instructions and the model is imitating those. Each has a different fix, and none is a longer prompt.
Should character details go in the description or the persona?
The persona. On Talaforge the description is the blurb people read while browsing and it never reaches the model, while the persona is what the character is asked to embody on every reply. Detail written into the description is read by humans and by nothing else.
What does temperature actually do to a roleplay scene?
It sets how willing the model is to pick a less obvious next word. Lower gives a consistent, predictable character; higher gives invention and, past a point, a character who contradicts details she established a page ago. It changes how sentences sound, never whether the scene has stakes.
Do I need a prompt to make the AI remember me?
No, and a prompt is the wrong tool: asking a character to remember something lasts as long as the conversation. Each character here keeps a memory you can open, correct, add to or delete, per character and per storyline, so a fact stated once does not need restating.
Can I bring prompts and characters from another platform?
Character cards in the common community formats import directly with persona, opening message and sampling parameters, and sampling presets import separately as JSON. Lorebooks and author's notes do not come across — if your setup depends on those, a front-end like SillyTavern remains the better fit.

Read next

Make a character in Telegram →