Promptchan Review: capable, but costly to choose (61/100)

Field	Value
Domain	promptchan.com
Category	AI image and video generation tools
Pricing	Freemium; $0–$0 shown in crawl
Pages crawled	20
Crawl date	2026-09-30

Promptchan has a usable visual and navigation foundation, but the buying case is underdeveloped. Usability scores 78/100, design execution 76/100, writing quality 73/100, and technical SEO 78/100. The homepage communicates the core offering with clear calls to action, and public pages are broadly crawlable.

The main weaknesses are concentrated in speed, messaging, and decision support. Performance scores 45/100. First impressions score 45/100 and audience and messaging 35/100 because the site does not clearly name its audience, cost, alternatives, or trust basis. Decision-support surfaces score 0/100, leaving visitors to infer differences between Plus, Premium, and Pro.

The priority is to publish clear audience and pricing guidance, then reduce mobile loading time.

Promptchan’s positioning is weak at 45/100 because the hero says “AI Girlfriend Chat, Video & Images” without naming a segment or giving a testable advantage. The site offers no adjacent proof beyond “11M users,” with no source, testimonials, or case studies in the supplied findings. Its free tier is not framed as a clear differentiator. Name the intended audience in the hero and navigation, replace the generic claim with a measurable promise, and support user-count claims with evidence. Keep the proposition tied to the actual image, video, and chat offering rather than relying on a category label alone.

Promptchan scores 35/100 for audience and messaging because visitors cannot quickly answer what it costs, why to choose it, or whether to trust it. The homepage CTA, “Get Started For Free,” provides no tier detail; the only price evidence is a $0 schema value. No comparison content, customer logos, case studies, or reviews were visible in the supplied findings. Add a visible pricing explanation, state the meaningful difference from alternatives, and show substantiated trust signals. Keep each claim specific: pricing must identify tiers and limits, while trust content should use named evidence rather than generalized reassurance.

Promptchan’s usability is strong at 78/100, with clear homepage calls to action and visually prominent content. The main task friction is information access: pricing appears in the footer under “Gems & upgrades,” while “Start creating free” leads to signup without clear tier guidance. Support also requires category browsing or search for common account questions. The “Explore AI Images” label does not clearly cover the site’s video and chat features. Move plan information closer to the primary CTA, link directly to common support answers, and use navigation labels that describe the full set of available functions.

Design execution scores 76/100. Promptchan has a consistent visual system, clear hierarchy, one H1 per page, and no skipped heading levels. Mobile checks found no horizontal scroll, generally sufficient tap targets, 16px inputs, and compensated sticky bars. The main defect is contrast: hero and FAQ body text measures 3.8:1 against white, below the 4.5:1 WCAG AA requirement for body text. Change the affected text to the supplied 4.54:1 color option or use a darker background. Also review the smaller mobile tap targets to preserve the otherwise sound interaction baseline.

Performance is weak at 45/100. The homepage records a 9.5-second LCP against the 2.5-second threshold and a 10.6-second interactive time on mobile. The supplied findings identify an oversized logo image, 13,087 image bytes, no lazy-loading attribute, and no preload; 34 blocking homepage requests also point to JavaScript execution cost. Prioritize the LCP asset and then reduce blocking script work through bundle analysis, code splitting, and tree-shaking. Treat these as measured lab findings, not a general claim about every device or visitor.

Writing quality scores 73/100 because the technical and commercial core is specific while the marketing funnel relies on vague language. The docs page gives an exact endpoint, POST /api/external/create, and auth headers; the upgrade page lists $14.99–$33.99 tiers and Gem allowances. By contrast, “Elevate your creative potential” and “Expand unlimited AI fantasies” add little information. The affiliate page delays its offer with “Dive into the world of limitless creativity...”. Rewrite those openings around the API function, benefits, and commission structure. Also give /signup, /gems, and policy pages distinct metadata and vary repetitive navigation anchors.

Decision support scores 0/100: the 20-page crawl found no pricing table, plan grid, or comparison route. Schema identifies Plus at $14.99, Premium at $23.99, and Pro at $33.99, but the site does not map those plans to users, trade-offs, or limits. Add a “Which plan is right for you?” layer to /upgrade with named use cases and explicit differences. Competitor comparison pages are a separate opportunity; they should use disclosed methodology and consistent axes rather than unsupported performance claims. The immediate conversion fix is a concise plan guide, not a larger feature list.

Risk and stability scores 75/100, with no evidence of sitewide noindex, robots blocking, or redirect loops. The main exposures are structural. /explore/ returns a second 200 beside /explore, /gems returns 200 while canonicalizing to /signup, and the API tutorial measures 484px scroll width at a 390px viewport. The homepage also has a 0.473 JS-only text share, creating render dependence for part of its content. Choose one slash policy, resolve /gems as either an alias or self-canonical page, constrain wide tutorial elements, and keep critical value propositions in server-rendered HTML.

Editorial QA scores 72/100. The strongest pages provide exact API details, pricing tiers, Gem allowances, and affiliate payout math: 50% of the first payment followed by 30% for four cycles. The weakest copy uses filler and generic commercial verbs, including “Elevate your creative potential,” “Expand unlimited AI fantasies,” and the affiliate opener “Dive into the world of limitless creativity...”. State the offer, API function, and commission structure in the first sentence of each relevant page. Repeated anchors—“explore,” “create,” and “chat” appear 38 times each—and duplicate metadata should also be corrected for clearer navigation and search presentation.

Technical SEO scores 78/100. The supplied evidence shows 200 responses for sampled public pages, robots.txt allowing crawling with a sitemap declaration, 21 sitemap URLs, a 404 guaranteed-nonexistent probe, HTTPS consolidation, and one-hop redirects from HTTP and www. The main cleanup is /explore/, which serves 200 without redirecting to /explore, plus /gems, whose canonical points to /signup. The API tutorial overflows at 390px, and eight /chat images lack alt attributes. Apply one slash policy, resolve the /gems intent, constrain wide content, add descriptive alt text, and retain critical homepage content in server-rendered HTML; no field CWV or GSC data was available.

Promptchan targets B2C users, but the site does not clearly name its audience. The strongest evidence is a practical interface: usability is 78/100, design execution is 76/100, and technical SEO is 78/100.

Three fixes should determine the next release. First, performance needs attention because mobile loading is slow. Second, the site must answer audience, cost, trust, and alternative questions; audience and messaging scores 35/100. Third, pricing needs actual decision support: Plus, Premium, and Pro are presented without guidance, producing a 0/100 score for that dimension. These are concrete communication and delivery issues, not a lack of product purpose.

This 13-dimension review is based on a crawl of 20 pages completed on 2026-09-30, plus the public dimension score table and site profile supplied to SiteList. The review covers positioning, messaging, usability, design, performance, writing, decision support, risk and stability, editorial QA, and technical SEO among the published dimensions.nnScores and findings are limited to the supplied crawl evidence. Read the full scoring approach in How SiteList scores.