Never run out of rizz.

Yes. Screenshots are uploaded over TLS, processed once, and auto-deleted from our storage within 24 hours. We never train models on your chats, never sell your data, and never share it with third parties. If you want a copy deleted sooner, email wingmanai.support@gmail.com and we'll wipe it within one business day.

Very accurate — and it keeps getting better. We use a two-pass pipeline: pass one runs high-fidelity OCR that preserves your Roman-script Hindi, Bangla and Tamil verbatim, then detects the exact code-switch ratio (e.g. 60% Hinglish, 40% English). Pass two writes replies matching that same mix. No stiff translations, no 'namaste bro' cringe.

That was the #1 thing we fixed in this build. Every reply is now grounded in what the other person actually said — you'll see a small caption like ↳ replying to: '…' under each suggestion. Anything the model can't quote from the transcript is filtered out server-side, so you don't get 15 fortune-cookie one-liners; you get 15 replies that fit YOUR conversation.

Yes — one tap in Settings → Billing. You keep Pro features until the end of your paid period, and any conversations, analyses and profile reviews you've saved stay in your account forever. No dark patterns, no 'call to cancel', no partial refunds trap. If you feel wronged, mail support and we'll make it right.

Anything you can take a screenshot of. Confirmed daily usage on WhatsApp, Instagram DMs, Bumble, Hinge, Tinder, Snapchat, Telegram, and Discord. If your screenshot is readable to a human, WingGuru can read it too — including grouped chats and reply-threads.

Absolutely. Nothing in WingGuru assumes your gender or the other person's. The tone chips (Playful, Bold, Chill, Respectful, Flirty) let you steer the vibe, and the AI mirrors YOUR voice from the transcript. Plenty of our top reviewers are women who use it to filter out low-effort openers and reply on their own terms.

Before writing a single reply, the AI extracts a structured transcript, detects the platform, the vibe (playful / flirty / dry / conflict / reconnect), the stage (opener / small talk / planning / post-date), and pulls out 'notable details' — running jokes, places you've mentioned, emojis you've used. All of that gets fed into a second model that writes replies grounded on those specifics. It's why 'she said she's from Kolkata' shows up naturally in your reply, instead of a generic line about coffee.

No. Your screenshots and transcripts are used exactly once — to produce your replies. They are never used to fine-tune or evaluate models, never seen by humans on our team, and never shared with third parties. Aggregated, non-identifying metrics (like 'average replies generated per week') are the only thing we track for product quality.

Free is a proper one-time taste: 1 chat analysis for life, 5 replies on it, and 1 profile review. Beginner (₹99 for 7 days) is a one-time starter pass for new accounts — 3 chat analyses per day, 8 replies each, 2 profile reviews, and 15 practice messages a day. Pro (₹299/mo or ₹2,399/yr) gives 7 chat analyses per day, 15 replies per analysis, 4 profile reviews per month, 60 practice messages a day, priority AI, and Deeper Insights — interest level, flirting score, mistakes you're making, and next-move suggestions. Elite (₹499/mo) steps up to 15 analyses per day, 10 profile reviews per month, 150 practice messages a day, plus a weekly personalised strategy report, VIP support and early access.

Email wingmanai.support@gmail.com. A real human (usually the founder) replies within 48 hours — most emails get answered same-day. Include a screenshot if you're reporting a bug. Refunds, plan changes, feature requests and 'this reply was weird' feedback all go to the same address.