DishGen scores 69/100 in this 13-dimension review, with clear recipe and meal-planning workflows as its strongest practical asset. Its most material weakness is design execution, where small tap targets and contrast findings reduce confidence in everyday use.
Reviewed by SiteList Engine · 11 of 13 dimensions · published Reviewed on October 21, 2026
Is this your site?
Claim itDishGen is a web product for AI recipe generation and personalized meal planning.
| Fact | Detail |
|---|---|
| Domain | dishgen.com |
| Category | AI recipe generation and personalized meal planning |
| Pricing | Freemium; DishGen Basic is free forever |
| Pages crawled | 38 |
| Crawl date | 2026-10-16 |
DishGen makes its core offer legible: it is an AI recipe and meal-planning product with a live prompt and a clear free-to-try path. The public score table shows its strongest listed dimensions in usability at 81.3/100, performance at 81.2/100, and technical SEO at 80.3/100. Accessibility is also strong at 76/100, while risk and stability scores 77.8/100.
The main drag is design execution at 37.2/100. The crawl reports 47 small tap targets and contrast findings, even though it found no measured horizontal overflow at 390px. Editorial QA scores 54.3/100, with repeated positioning and unsupported precision in product copy cited as concerns. Writing quality is 62.6/100, and the product leaves some audience and commitment details for visitors to infer.
The result is a useful starting point for home cooks and meal planners, with the clearest gains available in interface polish, publish-quality editorial checks, and sharper audience and commitment cues.
DishGen is immediately legible as an AI recipe and meal-planning product, but its primary cook remains implicit. The hero pairs "Create Custom Recipes & Meal Plans" with an above-the-fold prompt and "Free to try — no account or credit card needed," making the first action clear. The positioning is broader than its strongest use cases: leftovers, dietary adjustments, and week-long plans are more concrete than "Your Personalized AI Kitchen Copilot." The homepage also shows "OVER 1 MILLION AI RECIPES CREATED" beside WIRED, theguardian, CNN, and NPR marks without supplied titles, links, dates, or a definition of the count. Name the highest-value audience in the hero, then connect each press mark and explain what the million-recipe figure measures.
DishGen makes the first job easy to understand: enter ingredients, dietary needs, a recipe idea, or a meal-plan request and get a tailored result. The live hero prompt, Recipe and Meal Plan navigation, and "Free to try — no account or credit card needed" give visitors a low-friction start. The copy serves several groups—budget-conscious cooks, people reducing waste, dietary-goal planners, and anyone seeking inspiration—without prioritizing one. It also says users can target weight loss, muscle gain, or budget, but the supplied evidence does not explain recipe accuracy, nutrition limits, support, or cancellation detail. Add audience-led paths and a concise FAQ covering dietary requirements, plan contents, nutrition, support, and cancellation.
DishGen supports the core discovery tasks with prominent Recipe, Meal Plan, and Nutrition routes, visible account and trial calls to action, and a reachable contact form. The homepage exposes Create Account, Create Free Account, and Start 7-Day Free Trial, so visitors can see how to begin. The remaining usability gap is commitment clarity: the supplied crawl does not show the subsequent account or trial state, mobile parity, or confirmation behavior. Contact is a useful fallback with Email Address, Subject, Company, Message, and Send Message fields. State the next step and trial terms beside the entry buttons, verify the three primary paths on mobile, and provide a clear post-submit confirmation.
DishGen has a usable semantic foundation, but shared-template accessibility defects keep the result from being clean. Sampled pages expose navigation, main, and footer landmarks; no positive tabindex values were found; and the contact form explicitly labels Email Address, Subject, Company, and Message. The HTML lang attribute was null on six sampled templates, while PageSpeed flagged contrast and accessible-name issues. The homepage reports 81 unlabeled input candidates, but that count includes a display:none Facebook tracking form and needs an accessibility-tree check before all candidates are treated as visible failures. Create and mealplan place an H2 before the H1, recipes fails heading order, six sampled pages have one image needing an alt decision, and long pages have no skip link. Add the shared lang and skip link first, then recheck visible controls and reduced-motion behavior.
DishGen has a recognizable purple, yellow, and dark visual language, and its 390px capture showed no measured horizontal overflow. The execution breaks down in interaction details: the crawl reports 47 small tap targets, including a visible mobile menu button measuring 48x32px, below the 44px minimum height. PageSpeed also flags contrast, button naming, link distinction, link text, and form-label audits, although the supplied reports do not provide exact failing pairs or accessibility-tree nodes. The stylesheet shows substantial value drift: 366 colors, 32 font sizes, 30 radii, 44 spacing values, and 114 !important occurrences. Expand hit areas with padding, resolve the failing nodes directly, and consolidate the active palette, type, radius, and spacing tokens. Keep the consent banner from obscuring the final action or footer.
DishGen’s real-user interaction signal is good, but key mobile routes have a serious lab rendering problem. Homepage field data reports p75 INP of 130 ms, LCP of 2,549 ms, CLS of 0.12, and TTFB of 1,502 ms. The homepage lab run transferred 2,518 KB across 115 requests and measured 827.9 ms TBT; /recipes reached 9,053.6 ms LCP and /auth/signin reached 8,249.3 ms. Both route runs failed LCP request-discovery and render-blocking audits. Trace those exact LCP requests, make above-the-fold content discoverable in initial HTML, and defer non-critical JavaScript. The homepage’s 675 KB Facebook delivery and 197 KB Google Tag Manager delivery also warrant a third-party budget.
DishGen’s clearest writing tells a cook what to enter: "Simply enter your ingredients, dietary needs, a recipe idea, or meal plan request." The surrounding copy repeats broad positioning through "Your Personalized AI Kitchen Copilot," "The AI-Powered Kitchen Coach Built for YOU," and "revolutionary AI recipe generator" without adding a distinct proof point. Homepage text averages 38.6 words per sentence across 772 words, which makes the useful feature explanation harder to scan. Premium copy adds "25x more credits" and says most families can cover the cost through reduced food waste, but the supplied copy gives no nearby basis. Tag pages such as /tag/cheese also lack an H1 and meta description. Keep one concrete product metaphor, split long sentences, substantiate or narrow the premium claims, and add editorial framing to tag indexes.
This crawl identifies future visibility risks, not a measured traffic loss: Search Console and analytics were unavailable. Two sampled tag pages contain only 19 raw words versus 1,002 rendered words on /tag/grilled and 967 on /tag/coconut, with both marked JavaScript-dependent. If that pattern represents the template, discovery depends on successful rendering; server-render the headings, copy, and recipe links, then compare raw and rendered output after release. The HTTP root also takes two 308 hops before reaching the www HTTPS host, and the homepage page-facts artifact reports canonical:null. Consolidate the redirect and add a self-referencing homepage canonical. Homepage structured-data extraction was empty, so add only signals supported by visible content. The supplied 43 Wayback snapshots from 2021–2026 are context, not proof of stable traffic.
DishGen’s sampled pages have clear jobs, but the corpus still needs a consistent publish-quality gate. The homepage, create, mealplan, and premium surfaces repeat personalized or revolutionary kitchen-companion positioning instead of giving each page a distinct, evidenced promise. The tag corpus has a mechanical defect: /tag/cheese opens with the H2 "Popular Cheese Recipes" and the crawler records missing H1s and meta descriptions across many tag URLs. Premium’s "25x more credits" and reduced-food-waste savings claims also lack nearby comparison or calculation. Finally, the create and mealplan heading trees put an H2 before the H1. Add template checks for H1s, meta descriptions, and heading order; edit each funnel page around one user job; and substantiate or qualify the high-value premium claims. No schema-content mismatch is asserted because that evidence was not supplied.
The /premium and /create pages provide high-level marketing text rather than detailed usage or troubleshooting guidance. A probe for /llms.txt returned HTTP 404, so external machine-readable product guidance is also absent. Add an accessible FAQ on Premium and key interaction pages covering credits, plan contents, billing, and usage limits, while keeping Contact prominent in the footer. Publish an /llms.txt file only if its capability and tier summary can stay accurate.
DishGen has a functional technical baseline: the crawl reached 38 of 40 planned pages, verified a 21,782-URL sitemap, received 200 responses for the sampled public URLs, received a real 404 for the nonexistent probe, and found a valid HTTPS certificate. The main risk is the tag surface: /tag/grilled has 19 raw words versus 1,002 rendered words and /tag/coconut has 19 versus 967, with both marked JavaScript-dependent. Render their headings, explanatory copy, and recipe links in initial HTML. Consolidate the redirect, add and validate a self-canonical, and add structured data only where it matches visible content. Homepage field LCP is 2,549 ms and CLS is 0.12, so validate those signals after the technical changes.
DishGen is a usable, easy-to-try recipe and meal-planning product, but its 69/100 result is held down by design execution and editorial consistency.
The core experience is clear: visitors can reach Recipe, Meal Plan, and Nutrition routes, see account and trial calls to action, and use a contact form. Performance also scores 81.2/100, and usability scores 81.3/100.
The fix list is concrete. Address the 47 reported small tap targets and contrast findings, strengthen the publish-quality gate for product content, and make commitment details easier to understand. DishGen is best suited to home cooks and meal planners who want to turn ingredients, dietary needs, goals, budgets, or schedules into personalized recipes and meal plans.
This 13-dimension review combines a 38-page crawl dated 2026-10-16 with product facts and the supplied public dimension scores and summaries. The review covers positioning, audience, usability, accessibility, design, performance, writing, risk and stability, editorial QA, self-serve help, and technical SEO among the published dimensions.
Two dimensions were excluded as not applicable: 12 · Decision-support surfaces and 13 · Review-content integrity. Google Search Console and analytics access were unavailable, and no measured traffic drop was supplied; the risk and stability discussion is therefore a vulnerability pre-mortem rather than a traffic-loss diagnosis.
Read the full method at How SiteList scores. The byline describes this as a 13-dimension review.
DishGen generates custom recipes from ingredients, dietary needs, and goals, and supports personalized meal plans on the web.
Yes. DishGen Basic accounts are free forever. The supplied product facts confirm the free plan and do not provide paid-tier prices or billing cadence.
DishGen is aimed at home cooks and meal planners who want personalized recipes around ingredients, dietary needs, goals, budgets, or schedules.
Design execution is the clearest priority. The public score is 37.2/100, with 47 reported small tap targets and contrast findings that can make everyday use harder.