How to Set Up SpicyChat User Personas and Choose the Right Identity for Each Chat in 2026

Step 1: check capacity and protect your default profile

Prepare a neutral default and confirm space before creating either example.

The User Personas guide lists up to three personas by default: Get A Taste allows 10, True Supporter 50 and I'm All In 100. Check available capacity without deleting an established identity for this exercise.

Open Profile Settings through the top-right profile picture, following Profile Settings. Check Username, the unique identifier other users see; Name, the identity shown on your chat messages; Avatar, your chat image; and Highlights, the context characters receive. Use an invented nickname, a stylized nonidentifying avatar and a neutral preference for friendly fictional conversation. Remove addresses, employers, schools, contact details and personal history. Select Update at the bottom to save profile changes.

Default-profile Highlights are shared across conversations with any character. Do not treat persona selection as isolation or assume it hides shared profile information; these sources do not explain how conflicting profile and persona highlights are resolved.

Expected result: a saved nonidentifying default and room for two examples. Data minimization is not anonymity.

Step 2: create two concise fictional adult personas

Use distinct roles so you can recognize a wrong selection quickly.

Go to My Personas and choose Create New. Each persona has name, avatar and highlights; keep highlights to the recommended one or two sentences. Use these samples as field content:

Persona A template: Name: Mira. Highlights: 'Mira is a fictional 30-year-old museum curator meeting an adult colleague to plan an exhibition. She enjoys art and speaks only for herself.'

Persona B template: Name: Rowan. Highlights: 'Rowan is a fictional 34-year-old landscape designer meeting an adult colleague to plan a public garden. They enjoy plants and leave others' decisions to them.'

Avoid real faces in avatars; use clearly stylized, nonidentifying images. Complete the save action offered. Expected result: two distinguishable entries. Resolve saving failures before starting chats.

Step 3: choose the identity for each new chat

Match the persona to the scene before sending your opening message.

New character chats prompt you to choose a persona. Select Mira for an exhibition discussion and Rowan for a garden discussion. My Personas also supports adding, modifying, deleting and selecting default personas.

Choose by role, not list position. Keep any default persona nonidentifying too; that is separate from checking your default profile.

Expected result: the intended identity is selected. The guide does not establish mid-chat switching or promise to rewrite earlier messages. For a different role, start a separate conversation through the documented selection flow if you cannot verify a switch.

Step 4: verify selection with a harmless exchange

Check the displayed identity first; generated replies are only a consistency check.

The selected persona appears in the previous-conversations list. Inspect that listing before continuing. Then try one low-stakes prompt: 'Before we discuss the project, what name and role are you using for me?'

Expected result: the listing identifies your persona and the reply broadly matches the role. Replies cannot prove privacy, reliable memory or selection success; a confused reply alone does not prove selection failure.

If the listing is wrong, stop and check your selection. If only the reply is confused, restate the fictional name and task once. Never disclose your real identity to correct it.

Step 5: check prior-persona reuse with returning characters

Do not assume a new conversation starts with your general default.

A new conversation with a previously chatted-with character defaults to the persona used before. Check that preselection when returning to a familiar bot, especially when changing scenes.

After using Mira with an adult colleague, deliberately choose Rowan for a new garden scene with that character. That differs from continuing the exhibition conversation.

Expected result: recognizable conversations and identities. Stop if you cannot tell which you are about to use.

Step 6: understand deletion, downgrades and pricing

Review affected chats and billing before removing personas or changing plans.

Deleting a chat's persona makes that conversation revert to your default profile. After downgrading, if your persona count exceeds the new limit, the guide says to remove enough to fit the limit before you can 'create or edit any new ones.' It does not clarify whether editing existing personas is blocked.

Before deletion, note the affected fictional scenes and inspect your default profile again. Persona deletion is not documented here as transcript deletion or erasure of stored personal data; use the applicable privacy and deletion policy for those questions.

The September 2026 Pricing Update FAQ describes a September 8 change. Eligible older recurring subscriptions retain pricing while active and unchanged; changing tier or billing interval moves to current pricing, and expired subscriptions lose the old rate. Check your billing screen for currency, total charge, renewal and cancellation before proceeding. No account-specific price is verified here.

Expected result: an informed decision. Stop before confirmation if pricing, capacity or affected chats are unclear.

Finish safely: privacy, moderation and stop conditions

Personas organize identities; they do not exempt a conversation from platform rules.

The guidelines prohibit underage narratives, non-consensual scenarios, incest and family-related role-play, and hate speech. They also restrict real-person and photorealistic imagery. Violations can lead to suspension or termination. Keep both templates and chats within those rules; fictional names and private use do not create an exemption.

Do not put credentials, intimate identifying material or another person's private information in highlights or test messages. These four sources do not establish persona-level isolation, retention periods or confidentiality. Read current privacy terms separately before using sensitive material.

Stop for unresolved access, unclear charges, identity exposure or distressing replies. Report interface mismatches to official support with the platform, date and missing control, never passwords or full transcripts. Optional comparisons follow this help.

Frequently asked questions