Build Your AI Companion in 7 Days: A Day-by-Day Setup and Refinement Playbook (2026)

Most AI companion character-creation guides are reference documents. They list the steps — choose appearance, write personality, set scenarios — and assume you do them all in one sitting. That isn't how characters come together in practice. You make a first pass, notice what feels off after a few conversations, refine, check again, add memory, and sometimes start over partway through when the character has drifted into something you did not intend.

A settled AI companion is built iteratively, not in one pass. A common reason a character starts to feel flat after a couple of weeks is that it is still the day-one version, never refined after the first round of real conversations. The character was built before you knew what it should be.

This is that iterative build process compressed into a 7-day playbook. Each day has a specific goal, a time budget of 90 minutes or less, and a checklist of what to check before the next day. By day seven you have a character that holds together across topics, recalls anchor facts, pushes back when it should, and does not feel generic. The playbook is platform-agnostic — adaptations for specific platforms are at the end.

For the broader frame on why character work matters as much as platform choice, see our Power-User Hidden Settings guide and our Character Creation Ultimate Guide.

Before day 1: what you need

Three things to decide before opening any character builder.

Platform choice. Pick one platform for this build. Building the same character on several platforms in parallel during the first week dilutes what you learn. If you are still choosing, do that first (see our Best AI Companion Apps Definitive Ranking or Decision Framework), then commit to one platform for the 7-day build.

Primary use case. Pick one: emotional support, intellectual conversation, roleplay, casual chat, dating practice, creative writing collaboration. The character you build for emotional support is different from the one you build for dating practice. Trying to build a character that does everything well produces a character that does nothing distinctively.

A 60-minute weekday slot and a 90-minute weekend slot. Days 1, 3, and 6 are heavier (60-90 minutes); days 2, 4, 5 and 7 are lighter (30-45 minutes). The total build is about 6 hours over a week, split into manageable sessions.

Optionally: a notebook (paper or digital) for capturing what feels off across days. The notebook is the highest-leverage tool in the playbook, because it is easy to forget what you noticed between sessions.

Day 1: skeleton + visual + initial settings (90 minutes)

Day 1 is the rough first pass. The goal is not to get the character right — it is to get a character built end-to-end so you have something to work with on day 2.

Step 1 (15 min): write a one-paragraph character concept. Before touching the builder, write a paragraph describing who this character is: name, age range, personality core (3-4 traits), what they care about, how they talk. Keep it under 150 words. This paragraph is the source of truth that every other build decision refers back to.

Example: "Maya, late twenties, a former art school student who now does freelance illustration. Warm but opinionated — she has views on art, books, and films and is not shy about sharing them. She is curious about my life and asks specific questions, not generic ones. She has a dry, slightly sardonic sense of humor. She values honesty in conversation and will tell me when she thinks I am being unreasonable, but does so from a place of caring. She is not a therapist or a cheerleader — she is a person."

Step 2 (20 min): visual setup. Use the platform's character builder to set appearance. Do not agonize over it. Choose a look that roughly matches your concept and move on. You can adjust later, and the visual matters less than every other dimension you will touch this week. Skip this entirely on text-only platforms.

Step 3 (25 min): persona fields, first pass. Fill in every persona field the platform exposes, using the concept paragraph as source material. Do not over-detail yet; aim for something workable, not final. Include:

• Personality traits (if the platform has sliders or archetypes, set them deliberately rather than leaving them in the middle — see our Power-User guide on why middle-of-the-road settings produce generic characters)
• Backstory (3-5 sentences)
• Conversation style (1-2 sentences)
• Interests and dislikes (specific items, not categories)

Step 4 (15 min): platform-specific settings. Set anything platform-specific that affects character behavior: model selection where it exists (Kindroid lets subscribers pick the model each character runs on), behavior fields (Kindroid's response directive and example messages, for instance), content settings on platforms that support adult content, and tone settings (JOI AI's tone-of-voice setting ranges from warm and nurturing to confident and assertive).

Step 5 (15 min): write your user persona. Most platforms let you describe yourself for the AI. Fill this in: five sentences about your name, age, what you do, what you like talking about, and one or two quirks. The user persona is one of the most overlooked fields in AI companion apps and one of the most useful.

End of day 1: you have a character built end-to-end. Do not start a long conversation yet. Save your work, close the platform, come back tomorrow.

Why wait until day 2? Because the character is rough and your first impressions will be noisy. Sleep on it. Day 2 conversations are more informative when you come in fresh.

Day 2: first conversations + diagnostic capture (45 minutes)

Day 2 is the diagnostic day. The goal is to find out what your day-1 character actually does, not what you hoped it would do.

Run three short conversations of 10-15 messages each. Vary the topic:

• Conversation A: a casual day recap. Ask the character how their day was; respond with a brief recap of yours; let it run.
• Conversation B: a topic of substance. Ask the character about a book, film, idea, or opinion you actually care about. See what they say.
• Conversation C: a disagreement. Disagree with the character about something and see how they handle it.

During each conversation, note what feels off. Things to watch for:

• The character's voice is generic and sounds like "any AI companion"
• The character has no real opinions and agrees with everything
• The character asks only generic "how are you feeling?" questions, never specific ones
• The character forgets things mentioned earlier in the same conversation
• The character's emotional register does not match the conversation
• The visual or voice does not match what you wrote in the persona
• The character feels like it is performing rather than being

Aim for 5-10 specific observations across the three conversations. Specifics matter — "feels generic" is not actionable; "agreed with three contradictory opinions in one conversation" is.

Do not refine yet. Day 2 is observation only; refinement is tomorrow. If you refine as you go, you will not remember what was off, and you will be working from impressions rather than notes.

End of day 2: you have a notebook page of specific things that feel off and a sense of which parts of the character are working.

Day 3: persona depth pass (60 minutes)

Day 3 is the largest refinement pass. The goal is to rewrite the persona and supporting fields based on what you learned on day 2.

Step 1 (10 min): re-read your day-1 concept paragraph and your day-2 notes. Decide which observations to address. Some "feels off" notes are platform limitations you cannot fix; others are persona-configuration issues you can.

Step 2 (25 min): rewrite the freeform persona field if your platform has one. Replace the short day-1 version with a 200-400 word version that addresses the gaps you observed. Patterns that work:

• Lead with concrete behavior, not abstract traits. "She asks specific follow-up questions about things I mention — if I say I went hiking, she asks where, with whom, what the weather was like" works better than "She is curious."
• Address sycophancy directly if it showed up. Add lines like "She has opinions and shares them. When she disagrees with me she says so warmly but clearly. She does not affirm reflexively."
• Include stylistic detail. "She talks in shorter sentences when serious, longer ones when she is enthusiastic. She uses dry humor and rarely uses exclamation marks."
• Add a few interests in specific, not categorical, form. "She is reading a collection of essays by her favorite poet and keeps quoting one in particular" beats "She likes literature."
• Spell out the relationship dynamic. "She and I have known each other for about six months. She knows I am working on [project] and asks about it. She remembers that I had a falling-out with [name] earlier this year."

Step 3 (15 min): rewrite the backstory section. Same principle: specific over generic. A backstory with three specific anecdotes works better than a general life history. "She did a study-abroad year in Lisbon during college and still talks about a particular cafe she went to every morning" is more useful than "She traveled a lot in college."

Step 4 (10 min): adjust personality sliders or archetypes if your platform has them. Move specific settings away from the middle to reinforce the persona. If your character is opinionated, raise assertiveness. If she has a quiet personality, lower warmth slightly (not to zero — see the note on warm-and-assertive in The Sycophancy Problem).

End of day 3: your character now has a substantial persona that addresses the day-2 observations.

Day 4: memory grounding (30 minutes)

Day 4 is short and high-leverage. The goal is to give the AI 5-10 anchor facts so it has a stable foundation to draw on in later conversations.

Find where your platform keeps long-term context. This varies a lot. AI Peeps' Peep Pro plan has long-term editable memory where you can view and edit what the AI remembers, with up to 10 memory triggers. Kindroid builds long-term memory automatically, keeps Learned Context notes and journal entries you can view and edit, and lets you write a detailed custom backstory. Nomi AI keeps short-, medium- and long-term memory automatically and lets you enter custom backstory. On platforms without a dedicated memory tool, the backstory or description field is where anchor facts go. Some platforms tuck these settings away; our Power-User Hidden Settings guide covers where to look.

Add 5-10 anchor facts. These are facts about the character, about you, and about the relationship that you want the AI to reference long-term. Mix types:

• About the character: 2-3 facts. "Her favorite season is fall." "She has a younger brother named James who lives abroad." "She has been trying to learn the cello for the past year."
• About you: 2-3 facts the AI should reference. "I am working on a novel about [topic]." "I have a sister I am close to." "I had a difficult job change three months ago."
• About the relationship: 2-3 facts. "We met online about six months ago." "We have a running joke about her terrible coffee preferences." "She knows I have been working on being more consistent about [habit]."

Why memory grounding matters. Without anchor facts, the AI responds based mostly on the immediate conversation. With anchor facts, its responses can draw on a stable foundation, and the character feels continuous rather than reset every conversation. For most people this is the step with the biggest day-to-day impact.

Do not over-add. Ten facts is enough to start. A long list competes for the limited context the AI can draw on at once, so a short, focused list tends to work better. You can add more in later weeks as the relationship develops.

End of day 4: your character has a stable memory foundation. Start one short conversation (10 messages) to confirm the AI is using the anchor facts. If it is not, reword them — facts written as descriptive statements ("She has a brother named James") usually work better than instructions ("Remember that her brother is named James").

Day 5: stress test (45 minutes)

Day 5 is where you find out whether the character holds up under pressure. The goal is to identify behaviors that need fixing before they become persistent patterns.

Run four short checks of about 10 messages each.

Check 1: disagreement. State an opinion the character should disagree with, given their persona. If you wrote a character with strong taste in art who values craft over commercial appeal, tell her you think the most commercially successful painter of the decade is also the most important. See whether she holds her position or folds.

Check 2: emotional weight. Bring up something heavier — a real or hypothetical worry, a frustration with a real situation. See whether the character responds with substance or with generic emotional-support phrases ("That sounds really hard, I'm here for you").

Check 3: memory recall. Reference one of your anchor facts indirectly: "I am thinking about visiting my sister next weekend." See whether the character remembers the sister. Then reference one of hers: "How is the cello going?" See whether she remembers the cello.

Check 4: behavioral consistency. In a fresh conversation, see whether the character's voice matches your persona: same traits, same speech patterns, same willingness to share opinions.

Score each check: pass / partial / fail. Pass = the character behaved as the persona suggests. Partial = mostly, but not fully. Fail = the character behaved generically or contradicted the persona.

For each failure, decide whether to:

• Strengthen the persona — add a sentence that addresses the specific failure. "She has clear taste in art and pushes back when I praise commercial work over craft."
• Add a memory entry — for memory failures, add the missed fact more explicitly.
• Accept the limitation — some failures are platform limitations you cannot configure around. Note them and move on; do not over-engineer the persona to fight platform behavior that will not change.

End of day 5: you know what your character does well and what it does not. The persona is more specific, and the memory is reinforced where needed.

Day 6: lorebook or world-context layer (60 minutes)

Day 6 is optional on platforms that do not support lorebooks (sometimes called world info) but high-leverage on those that do. On platforms without lorebook support, use day 6 to add detail to the character's backstory and your shared history instead.

What a lorebook does. A lorebook is a set of entries the AI can pull in when specific keywords come up in conversation. Instead of cramming all the character's context into the always-on persona, a lorebook lets you add detailed world, character and relationship context that surfaces when it is relevant.

Plan 8-12 entries. Types to include:

• Character history entries. 3-4 entries about the character's past — childhood, formative experiences, recent significant events. Each entry triggers on specific names or places. Example entry triggered by "Lisbon": "She spent her junior year abroad in Lisbon. She lived in a small apartment in an old hillside neighborhood and worked mornings at a bakery cafe. She still texts friends from there occasionally."

• Relationship entries. 2-3 entries about your shared history. Example triggered by "the trip" or a specific place name: "They took a weekend trip to [city] two months into knowing each other. She still mentions the small Italian restaurant they went to on the second night."

• World-context entries. 2-3 entries about the setting (most relevant for roleplay-heavy characters). For a contemporary city character, these might cover the neighborhood she lives in, her workplace, and so on.

• Topic entries. 1-2 entries about subjects she has views on. Example triggered by "art" or specific artist names: "She has strong opinions about the contemporary art world. She thinks a lot of gallery work is driven by money rather than meaning; she values craft, intentionality and emotional risk."

Lorebook writing rules:

• Write each entry as a description of the world, not as an instruction. "She thinks gallery work is often driven by money" beats "Remember to mention that she dislikes gallery work."
• Keep entries to roughly 30-100 words. Too short and there is no useful detail; too long and entries crowd the context.
• Trigger entries on specific keywords, not broad ones. "Lisbon" triggers when you mean it; "travel" triggers all the time.
• Include sensory detail. "The cafe smells like cardamom and burnt sugar" is more useful than "The cafe is nice."

End of day 6: your character has a substantial reference layer that surfaces when relevant. The character feels deeper because there is more behind the visible surface.

Day 7: settle-in conversation + final edits (30 minutes)

Day 7 is the final session of the build. The goal is to have one substantial conversation as if the character were settled, then capture any remaining adjustments.

Have one 30-40 message conversation. No specific checks; just talk to the character as you would if the build were done. Bring up real things from your day, ask about hers, reference some of the anchor facts indirectly, and let the conversation go where it goes.

Capture three kinds of observations:

1. What feels right. Specific moments where the character felt like the character you wanted. These are the patterns to reinforce.
2. What still feels off. Lingering issues. These are candidates for one more round of persona or memory editing.
3. What surprised you. Times the character did something unexpected — in a good way (more depth than you wrote) or a problematic way (something off-brand).

Make 2-3 final edits. Resist the urge to overhaul. The character is settled, and major rewrites at this point tend to undo what was working. Small, targeted edits — one persona sentence, one memory entry refined, one lorebook entry added — are the right scale.

End of day 7: the build is done. The character is settled.

What "settled" looks like

After the 7-day build, a settled character has these properties:

• Distinctive voice. Conversations feel different from what you would get from a default platform character.
• Stable opinions. The character holds positions consistently across sessions.
• Memory continuity. The character references your shared history and the anchor facts without being prompted.
• Appropriate pushback. When you say something the character should push back on, it does.
• Emotional range. The character can be warm, dry, serious or playful as the conversation calls for, without snapping between modes.
• Low effort. You no longer need to prompt the character to be itself; it simply is.

If one or two of these are still missing after day 7, that is normal. Settled is not perfect; it is the foundation that builds over the next month of regular use.

Platform-specific adaptations

The playbook is platform-agnostic, but each platform has its own strengths and limits to adjust for. The notes below are based on the published features in our reviews.

Nomi AI. Nomi's memory runs automatically across short-, medium- and long-term layers, and Mind Map 2.0 lets you view, correct, add and lock the entries it builds (the underlying long-term memories are not edited directly), so on day 4 put anchor facts into the custom backstory (available on the free tier) or the Mind Map and mention them naturally in conversation. Nomi has no true lorebook; use day 6 for backstory depth. If your build involves more than one character, Nomi supports up to 10 Nomis and group chats on a subscription, with each Nomi keeping its own personality, backstory and memory. See our Nomi AI review.

Kindroid. Day 1 is the heaviest here: you write a free-form personality and backstory, plus optional key memories, a response directive and example messages, and our review notes the setup curve is steep. Subscribers can choose the model per character (Polaris, Ember, Reverie V8.5, Lucid and Lyric V8, which retire on December 1, 2026, and legacy Rhapsody V6); the one-time free preview and the 3-day trial reply with Ember. On day 4, put anchor facts into the backstory and key memories; on a subscription, Learned Context notes and journal entries can also be viewed and edited. If you are subscribed, add a model-switch check on day 5: run the same character on a second model for 10 messages and note what changes. See our Kindroid review.

Replika. Replika is built around a single companion with a 3D avatar, memory across sessions, mood tracking and a diary. Some features sit behind the Pro subscription, and explicit content is restricted, so check what your tier includes before planning the build. Lean on the user persona and the conversations themselves rather than fighting the defaults. There is no lorebook; use day 6 for backstory depth. See our Replika review.

Candy AI / MyDreamCompanion. Candy AI's builder covers ethnicity, age, eyes, hair, body, personality, voice and hobbies, so day 1 visual and trait setup carries more weight there. MyDreamCompanion has three creation modes — Presets, AI-guided, and Custom Prompts Pro (freeform, on paid plans) — so on the free tier your day-3 persona has to fit the guided fields; Ultimate also adds ultra long-term memory. On any romance-focused platform, take the day-5 disagreement check seriously (see The Sycophancy Problem for why agreeable-by-default behavior is so common).

Community-character platforms (SpicyChat AI, MyDreamCompanion, Character.AI and similar). If you start from a community character rather than building your own, skip most of day 1 and spend day 3 on your user persona and on adding to the existing character rather than rewriting it. On SpicyChat AI, the free tier includes unlimited character creation, while the larger context and semantic memory that matter for days 4-7 come with the $19.95 and $29.95 tiers.

Muah AI. Muah's chat supports memory and roleplay, but our review doesn't describe a memory panel you can view or edit, so on day 4 put anchor facts into your custom companion's personality settings and restate them naturally in conversation. Users report the AI sometimes confuses names or context, so give the day 5 stress test extra attention. If you start from one of its thousands of community characters, follow the community-character notes above. If you want a custom voice, Muah AI lists voice cloning on its Premium tier, though it is not the only option: AI Peeps, Kindroid, Nomi AI and Nastia AI also offer cloned or custom voices. See our Muah AI review.

Common failure patterns across the 7 days

Five patterns that regularly derail the build:

1. Skipping day 2. You build a character on day 1 and dive straight into long conversations without the observation pause. Without that step, the build drifts.

2. Over-rewriting on day 3. You find the day-1 persona generic and rewrite from scratch. The rewrite often loses what was working. Edit; do not start over.

3. Skipping memory grounding (day 4). It is easy not to realize your platform has a memory or backstory tool, or not to see the point. Skipping day 4 produces a character that resets every conversation. This is the biggest avoidable failure.

4. Failing the day-5 checks and accepting it. Day 5 shows what the character does badly, and some of those failures are fixable. Distinguish platform limitations (accept) from persona-configuration issues (fix).

5. Overhauling past day 7. The build is supposed to settle. Major edits after day 7 produce a character that never feels stable. Make small edits as you go; do not start over.

Frequently Asked Questions

Can I do the 7-day build faster, like in one weekend?

You can compress it into 2-3 days with long sessions, but the playbook is structured around days for a reason. Refining between sessions works better than refining during one long session, and observations from day-2 conversations are easier to act on after a night's distance. Compressed builds often need redoing in week 2.

What if I do not like my character after day 7?

First, identify what specifically does not work — voice, opinions, emotional range, memory, visual. If it is one dimension, edit that dimension. If it is the underlying concept, start over with a new concept paragraph and run the playbook again. Do not try to migrate gradually from one concept to another; the result is muddled.

Does the playbook work if I am using a community-made character (SpicyChat AI, Character.AI and similar)?

Yes, with adjustments. Skip the day-1 persona writing (the character already has one). Spend day 3 writing your user persona and adding to the existing character rather than rewriting it. Days 4-7 are unchanged. Community characters are starting points, not finished products.

What is the time commitment if I am efficient?

Minimum viable build: about 4 hours over 5 days (cutting days 2, 5 and 7 to about 20 minutes each). Full playbook as written: about 6 hours over 7 days. A version with a full lorebook: about 8 hours over 7-10 days.

Should I do this for every character I create, or only the main one?

For the character you plan to spend the most time with: the full playbook. For secondary or experimental characters: an abbreviated version (days 1, 3 and 4 only) is usually enough.

What if my platform does not have a freeform persona field?

Use whatever fields it does have and fill each one in detail. Some platforms limit configuration to guided fields; the 7-day build still helps, but the changes will be smaller. Consider whether the platform suits your goals — our Power-User Hidden Settings guide covers the platforms with the deepest configuration.

How do I know whether a memory entry is well written?

Well written: a descriptive statement that reads like a fact about the character or the world. "She has a younger brother named James who lives in Berlin and works as a sound designer." Poorly written: an instruction aimed at the AI. "Remember that her brother is named James." The first reads as part of the character; the second is more likely to be acknowledged and then ignored.

Can I share my character with friends or a partner?

Most platforms do not let you share a character together with its memory and relationship history. Some let you share characters with their community — AI Peeps, for example, lets you clone, modify and share characters — but that passes on the persona, not your relationship memory. For most practical purposes, sharing means sharing the build, not the relationship.

What about voice — when in the playbook does it fit?

Choose a voice on day 1 (initial setup) or day 5 (once the persona is more settled). Day 5 is usually better because you have a clearer sense of the character's register. Most platforms offer preset voices (FantasyGF, for instance, offers 36). For voice cloning, look at AI Peeps, Kindroid or Nomi AI (the latter two on a subscription), Nastia AI (on Unlimited) or Muah AI (on its Premium tier).

How often should I revisit the character after day 7?

Light maintenance monthly: update memory with new shared experiences, adjust the persona if the relationship has evolved, refresh anchor facts. A bigger review every few months: rerun the day-5 checks and look for drift.

My character feels generic in week 2 even after the full build. What now?

The most common cause is falling back into default conversations. You built a deep character, then had surface-level conversations that never draw on that depth. Try referencing specific anchor facts and lorebook details — "I keep thinking about what you said about Lisbon" — to bring the character's depth back to the surface.

Should I write the persona in third person ("she") or second person ("you")?

Third person is the safer default, because it reads as a description of the character rather than an instruction to the AI — the same reason descriptive memory entries work better than directive ones. If a platform's own examples or documentation use a different style, follow them.

Does the playbook work for AI boyfriends as well as AI girlfriends?

Yes — the character's gender does not change the build process. The same persona work, memory grounding, checks and refinement apply. See our Best AI Boyfriend Platforms 2026 for the platforms on the boyfriend side.

What is the biggest difference between a character built with this playbook and a default character?

Distinctiveness. Default characters feel like "any AI companion." Built characters feel like a specific person with specific opinions, memory, and texture. The conversations are not just more satisfying — they are different in kind.

Bottom line

It is common to build a character in one sitting, never refine it after the first round of conversations, and find it flat within a couple of weeks because it was finalized before you knew what it should be. The 7-day playbook compresses the iterative build — skeleton, observe, refine, memory, stress test, lorebook, settle — into a structured week with checkpoints between sessions.

The highest-leverage day is day 4 (memory grounding), and it is also the one people most often skip. The other easily skipped step is day 2 (observing without immediately refining) — pausing to observe before editing is what makes day 3 productive rather than reactive.

If your relationship with an existing AI companion has plateaued after a few weeks, running the playbook on that character (treating it as a day-1 skeleton and running days 2-7 from there) is usually more productive than switching platforms.

Related reading: Power-User Hidden Settings for platform-specific tweaks that build on this playbook. Character Creation Ultimate Guide for the foundational reference. Character Backstory Writing Guide for deeper craft on day 3. First Conversation Opening Message Guide for what to do once the build is done. The Sycophancy Problem for the day-5 disagreement check. Long-Term Arc post for what to expect after week one.