Nextpart AI positions itself as a character-first chat platform where people create, customise and converse with AI characters — including optional voice responses and in-chat image generation. The service markets a permissive, adult-oriented approach to roleplay while also offering public and private character publishing, a marketplace of user-created characters and a freemium + credits + subscription model for media features and larger memory allowances. For more companion options and category-level comparisons see the AI girlfriend category on SteveJabba.
This review summarises what Nextpart officially advertises and what its public documentation reveals as of the August 2026 evidence snapshot: available features, how media and credits are metered, moderation and privacy trade-offs, and practical purchase cautions where the site lacks transparent pricing or consistent credit accounting.
- Nextpart.ai — Homepage
- Nextpart.ai — FAQ
- Nextpart.ai — Privacy Policy (Last Updated Feb 5, 2024)
- Individual user reports are not independently proven facts.
See more AI girlfriend reviews on SteveJabba.
Quick answers — Should you consider Nextpart AI?
- Best for: Users who prioritise character design, voice replies and on-the-fly image generation and who want few built-in content filters for mature roleplay.
- Not for: Buyers who need clear, public pricing, guaranteed data deletion or a strong mainstream trust profile (app‑store listing and large third‑party review corpus are limited).
- Core strengths: Flexible character creation, contextual image generation in chat, voice synthesis, and marketplace browsing for user-designed characters.
- Main concerns: Inconsistent public statements about credit costs, incomplete price visibility on the open Pricing page, explicit privacy language allowing content reuse for model training, and mixed trust signals from independent site-safety checks.
- SteveJabba score: 3.7 / 5 — a mixed-to-good fit with notable trade-offs (see full verdict).
What Nextpart actually offers
Nextpart’s public product pages and blog describe a suite focused on conversational roleplay with media: a character builder (appearance, personality and backstory), text chat, paid voice responses and in-chat image generation that produces character headshots tied to conversation context. A pricing page and FAQ show a credit-metering model plus subscription tiers that add monthly credits and larger “memory” buckets (the site cites things like 16k/32k memory per tier as selling points). The company name on legal pages is Envision Investment LLC.
Two important clarifications come directly from the site’s documentation: guest/anonymous users can try the system with a limited sentence or credit allowance, and users can publish characters to a community marketplace or keep them private. The blog and guides also document headshot generation workflows and credit costs in examples (these blog examples are promotional guidance rather than a formal, public price list).
Chat quality, memory and continuity
Text chat is the product core. Nextpart advertises “human-like interactions” and roleplay use cases; it also promotes memory as a feature for longer, more immersive conversations. Memory allotments are explicitly listed on the Pricing page as a differentiator (larger tiers promise 16k or 32k memory), and the FAQ hints at character onboarding behaviours (for example, a new public character needs a number of sentences before it becomes visible to others).
What the public pages do not provide are technical provenance details for the underlying models (no named LLM vendors or model families are published) and independent benchmarks of conversational quality. Community posts and small forums mention mixed experiences with session continuity: some users report smooth context retention, others say memory can be inconsistent across long interactions. Treat the memory claims as product marketing with partial independent confirmation.
Voice, images and beta video
Nextpart emphasises media features more than many niche companions. Voice responses are an advertised deluxe/premium feature; product pages and the FAQ describe voice synthesis for characters and label it as a paid enhancement. In-chat image generation (especially headshot/character images derived from chat context) is documented in guides and blog posts. The Pricing page also lists video generation as a beta feature for certain tiers, but there are no public technical specifications or sample galleries on the public pages to judge quality.
Operationally, the site treats these media features as metered: blog posts and guides show per-headshot credit examples (headshot generation cited in guidance), and the Pricing/FAQ pages confirm credits cover chat and image generation. However, the public material is inconsistent about exact costs per message or per-conversation unit (see Pricing and Credits below).
Pricing, credits and what’s unclear
The business model is freemium: free guest access with limited sentences/credits, paid credits packs and subscription tiers that bundle monthly credits and memory. Payment methods listed include Stripe, PayPal and a crypto payment option. But there are a few practical obstacles for buyers relying on the public site:
- The Pricing page displays tier names and feature bullets but does not show definitive dollar amounts on the open page — price fields appear blank ($0.00 placeholders), implying the full purchase flow or account sign-in is required to view prices.
- The public documentation contains conflicting statements about credit consumption: the FAQ says “Each text response in chat sessions consumes 10 credits” while the Pricing UI elsewhere uses phrasing like “20 credits for each AI-assisted conversation”. That inconsistency means you cannot reliably compute dollar-per-message without logging in.
- Blog examples and guides show trial credit offers and per-headshot costs in illustrative terms (for instance a headshot example quoted at X credits in blog text) but these are guidance numbers, not a verified price table.
Bottom line: if precise per‑message or subscription dollar pricing matters to you, expect to sign up for an account or contact support before purchasing. The publicly visible pages do not provide the full, consistent price breakdown needed for a buying decision.
Content rules, moderation and category signals
Nextpart publicly positions itself as permissive for mature roleplay and markets a “no NSFW filter” stance in marketing text; at the same time its legal pages and FAQ impose boundaries: users must be 18+, and content involving minors, sexual exploitation, sexual violence, bestiality and similar illegal/abusive categories is prohibited. The Terms also restrict misuse of voice features (e.g., uploading third‑party voice recordings without consent and impersonation/deepfake abuse).
There is a practical moderation ambiguity to flag: the public site indexes many adult categories (some labels, including controversial ones, appear in category tags), and the Terms do not categorically ban every adult roleplay type that appears in the UI. The company states it conducts reviews and will remove policy‑violating material and provides reporting contacts, but the combination of aggressive marketing and a permissive UI taxonomy means buyers should expect a community where mature roleplay is common and moderation boundaries may be enforced after the fact rather than proactively.
Privacy, data use and deletion caveats
Nextpart’s privacy policy (last updated February 5, 2024) documents routine data collection: account details, payment metadata, chat content, images and device/usage data. Importantly, the Privacy and Terms both state the company may use user content to operate and to train its machine learning models and that some content or character states may be retained even if a user deletes them (the policy includes an explicit exception for popular/public characters). Those are normal commercial choices for many AI services but they represent a trade-off: creators should not assume their chats or generations are removed from training datasets or from backups.
Support and abuse reporting email addresses are published on the site; users can request deletion of chat sessions and characters, but the public documentation contains caveats and does not promise wholesale removal in all cases. The site also lacks a clearly visible, formal refund policy on the public Terms or Pricing pages.
Trust signals and availability
Nextpart has an active public site, blog and FAQ, but independent trust signals are mixed. Niche AI tool directories list Nextpart as a known option; some automated site-safety scanners and smaller reputation sites flag low or suspicious trust scores. There is limited presence in mainstream review platforms and no clear, verifiable Apple/Google Play store listing tied to the brand in public searches — the blog’s references to mobile support are therefore ambiguous and should be verified if you prefer a native app experience.
These factors matter if you plan to make recurring payments: confirm payment methods, refund mechanics and verify support responsiveness before committing significant funds.
Who should use Nextpart — and who should avoid it
- Good match: People who want a character marketplace, custom character creation, voice-enabled replies and in-chat image generation and who are comfortable with a platform that allows mature roleplay and uses content for model improvement.
- Consider alternatives if you: require public, fully transparent pricing before signup; demand strict guarantees that uploaded content won’t be used for model training; or need a product with a long history of mainstream app‑store reviews and strong third‑party trust signals.
Practical alternatives
If Nextpart’s combination of features appeals but you want different trade-offs, consider evaluating:
- Companion/chat platforms that publish transparent pricing and data policies (look for named model provenance and clear content-retention/deletion guarantees).
- Specialised image/voice generator products paired with a separate chat client, if you want more control over media ownership and export formats.
- Mainstream conversational assistants if you prioritise strong privacy controls, app-store distribution and a larger independent review corpus.
FAQ
Does Nextpart allow mature roleplay? Yes — the public marketing positions the platform to support mature content, but the Terms and FAQ prohibit illegal or abusive categories such as minors, sexual violence and bestiality. Moderation actions are possible and reporting channels are provided.
Are prices visible without an account? Not fully. The public Pricing page lists feature bullets and memory/credits per tier, but visible dollar amounts are not shown openly; the site appears to require sign-in or a purchase flow to reveal the final prices.
Will my chats be used to train models? The Privacy policy states that user content may be used to operate and to train machine learning models; deletion requests are possible but the policy includes retention exceptions for legal compliance and for public/popular characters.
Is voice and video available? Voice replies are advertised as a deluxe/premium feature and in-chat image generation is documented; video generation is listed as a beta offering but public technical details and sample outputs are not provided on the public pages.
Final verdict
Nextpart AI is a distinctive, feature-rich companion platform for character-driven chat: its strengths are character creation and sharing, voice replies and contextual image generation, plus marketplace-style browsing for user-created personalities. Those strengths make it an attractive option if you prioritise creative roleplay tools and media features.
However, the public site leaves important buying questions unresolved: inconsistent credit statements across FAQ and Pricing, no open dollar pricing on the public Pricing page, and privacy terms that allow content to be used for model training and to be retained in some cases. Trust indicators are mixed outside niche tool lists, and app availability is unclear in public app stores. These are material considerations if you intend to pay for credits or a subscription.
Score rationale: a 3.7 reflects helpful functionality and interesting media capabilities, offset by transparency, privacy and trust trade-offs that reduce its recommendability for users who need firm guarantees before spending money. If you value creative character play with voice and images and accept the documented privacy and pricing caveats, Nextpart is worth testing via its free/guest tier; if you need full pricing visibility, stronger deletion guarantees or mainstream app-store presence, look instead at alternatives with clearer public policies.
Disclosure: Some links on this page may be affiliate links. SteveJabba.com may earn a commission if you join through them, at no extra cost to you.