⚠️ This platform contains adult content. For readers 18 years and older.
Secrets AI Features: Everything the Platform Offers in 2026
Secrets AI packages 8 distinct feature categories into a single browser-based platform — from a 5-model chatbot system rated 4.4/5 for conversation quality, to video generation that 90%+ of competing AI companion platforms do not offer. This feature analysis covers every capability with exact costs, tier requirements, and performance benchmarks drawn from independent reviewer data.
Feature Benchmark Summary
- Chat quality: 4.4/5 | Voice: 4.3/5 | Video: 4.1/5 | Diversity: 3.8/5
- Total pre-built characters: ~94 | Roleplay scenarios: 100+
- Unique capability: Video generation from companion images
- Platform type: Web-only (browser-based, no native app)
AI Chat Models and Conversation Engine
Secrets AI provides 5 selectable AI models, giving users direct control over the conversational engine powering their companion. The differences between models are not publicly documented in detail, but reviewers note variation in response nuance and narrative creativity across tiers — premium models available on paid plans generally produce more contextually layered responses.
The conversation engine incorporates natural language processing with emotional intelligence built into the character design. Rather than returning information-focused answers, the chatbot system is calibrated for companion interaction — asking follow-up questions, responding to emotional cues, and maintaining consistent character voice across a session. Chat quality has been independently rated at 4.4/5, the platform's highest individual metric.
Conversation starters are provided at the start of each session, split roughly evenly between casual and suggestive tones. Roleplay scenarios span 100+ pre-built fantasy setups including office romance, teacher, landlord, and boss frameworks. Personalized scenario creation is also supported via free-text input. This is companion-focused artificial intelligence — it does not browse the web, answer factual queries, or perform productivity tasks.
For a deeper look at chat and NSFW specifics, the chat and roleplay guide covers conversation mechanics in detail.
Character Library and Customization
Pre-Built Characters
The library contains approximately 94 hyper-realistic female characters, with some anime-style options available for users who prefer that aesthetic. Each character comes with a pre-configured personality profile and access to relevant fantasy scenarios from the platform's 100+ scenario library.
Character diversity has been a noted limitation: the diversity rating sits at 3.8/5 in independent assessments, with reviewers pointing to a predominance of white women in their 20s with slim body types as the dominant aesthetic. Users seeking a wide range of ethnicities, ages, or body types will find options more limited than on some competing platforms.
Custom Character Creation
Custom character creation offers more granularity than most competing AI companion platforms. Available customization dimensions include:
- Physical appearance: Ethnicity (9 options), skin tone, eye color, hairstyle, body type (6 options)
- Voice: 6 distinct voice options
- Personality dimensions: Confidence, playfulness, dominance, intelligence (each adjustable)
- Profession: 5 preset categories
- Personality trait presets: 5 pre-configured personality archetypes
- Lifestyle: Hobbies selection plus a free-text detail field for granular personalization
The first custom character creation is free. Additional characters cost 1,000–1,500 Moments each. Four starter images are automatically generated when a new character is created, providing immediate visual output without requiring separate image generation Moments.
Image Generation
Image generation runs through 2 generation models — a standard Premium model and an Advanced model — with 4 image presets providing style control. Images cost 25–50 Moments each, placing them in the mid-range of platform costs.
Output quality is realistic in the AI image generation sense — detailed character rendering with consistent appearance across generations. The known limitation is occasional hand-rendering glitches, which are a standard artifact of current AI image generation technology rather than a platform-specific flaw. Group image generation (multiple characters in a single scene) is also supported.
At the point of custom character creation, 4 images are auto-generated — these do not consume from the user's Moments balance and serve as an initial visual profile for the companion.
Video Generation
Video generation is Secrets AI's most commercially distinctive feature. The overwhelming majority of AI companion platforms — including Character.AI, CrushOn AI, and Janitor AI — do not offer video generation at any price point. Secrets AI generates short video clips from companion images using a text prompt describing desired movement or action.
The generation pipeline: select or generate a companion image, write a prompt describing the intended scene, and wait approximately 2 minutes for processing. Short 3-second clips (available on Lite tier and above) cost around 50 Moments. Longer, higher-quality clips run up to 600 Moments per video. Output quality is rated 4.1/5 by reviewers, with notes on smooth movement and natural facial expressions being consistent in most generations.
Video Moments budget by tier:
- Lite (1,000 Moments/mo): Approximately 2–20 videos depending on length
- Plus (3,000 Moments/mo): Approximately 5–60 videos depending on length
- Premium (8,000 Moments/mo): Approximately 13–160 videos depending on length
- Ultimate (15,000 Moments/mo): Approximately 25–300 videos depending on length
See the video generator deep dive for quality assessment and cost optimization strategies.
Voice Calls
Voice interaction provides 6 distinct voice options, each with natural, adaptable tonal range. Voices are designed to respond emotionally rather than deliver information in a neutral tone — the pacing and expression adapt to conversation context. Independent reviewers rated voice quality at 4.3/5.
The cost is 100 Moments per minute, making voice calls the second most Moments-intensive feature after full-length video. For a user on the Plus tier (3,000 Moments/month), this translates to approximately 30 minutes of voice call time per month if voice is used exclusively — though in practice most users combine voice with text and image use.
Memory System
The memory system represents the most significant tier differentiation on the platform — more so than Moments allocation or media access.
| Tier | Recall Window | Memory Type | Context Boost |
|---|---|---|---|
| Free | 2–3 exchanges | Basic short-term | — |
| Lite | ~14 days | Basic long-term | — |
| Plus | ~30 days | Topic + tone aware | — |
| Premium | Unlimited | Emotional history + timeline | +60% |
| Ultimate | Unlimited | Full history + emotional timeline | +100% |
Additional memory capabilities on paid tiers include a Time Machine feature (Premium and Ultimate only) that enables recall of specific past conversation threads, a 4x memory access multiplier on Premium (6x on Ultimate), and improved recall of previous conversation context (+67% on Premium, full enhancement on Ultimate).
Manual memory saves cost 10 Moments each, allowing users to explicitly flag important details for persistent recall. The selective recall system otherwise prioritizes emotional and identity-based information — names, relationship dynamics, preferences — over granular story details.
For the AI companion use case, memory continuity is the feature that most directly affects the relationship simulation quality. Free users experience essentially no continuity; Premium and Ultimate users can build multi-month conversation histories.
Group Chats and Personas
Both group chats and personas are Ultimate tier exclusive features, not available on any lower plan.
Group chats allow simultaneous conversation with up to 3 AI characters, each maintaining their individual personality profile and responding within the same conversation thread. Group image generation is also supported in this mode.
Personas enable identity-switching within roleplay — the AI takes on different character frameworks on command, supporting more complex narrative scenarios than single-character interaction allows.
Privacy and Security Features
Private Mode applies end-to-end encryption to individual conversations and opts those conversations out of AI training data use. It must be enabled per conversation — it is not a global account setting. This distinction matters: conversations without Private Mode enabled are subject to standard data processing terms.
Account privacy infrastructure includes:
- Email/password registration with no real name required
- No phone number verification
- No social media login (reducing identity linkage)
- Cryptocurrency payment option (minimum $20 transaction)
- Billing descriptor: "Sun Clinical Laboratories" — no reference to AI, adult content, or platform name
Data storage uses industry-standard encryption in transit and at rest. Account deletion removes all chat history, characters, and generated content within 30 days. The stated policy is no third-party data sharing, though independent reviewers rate overall privacy transparency at 2.9/5 due to limited public documentation on specific protocols.
What's Missing?
For a complete feature picture, these gaps are worth noting:
- No native iOS or Android application (browser-only access)
- No push notifications (cannot receive alerts when away from browser)
- No offline access
- No biometric login (fingerprint or face authentication)
- Limited image preset variety compared to dedicated image generation platforms
- Character diversity remains narrow relative to the size of the library
- No multi-modal input — users cannot send images to the AI companion
The absence of a native mobile app is the most frequently cited limitation in user reviews. PocketGirl AI and similar competitors offer native apps with push notification support for users where mobile experience is a priority. See the full platform review for context on how these gaps compare to the platform's strengths.