Warbl
Boost engagement and grow your audience effortlessly on X
191 followers
Boost engagement and grow your audience effortlessly on X
191 followers
Generate perfect replies on X (Twitter) in one click. AI-powered Chrome extension with personalized voice profiles, smart web search, and image analysis.
Free Options
Launch Team
CloudflareStill paying to deploy your apps? Try Cloudflare Workers for free
Promoted
@0x_alx Does it analyze our past tweets to learn the syntax or we are just selecting a generic "witty" or "professional" preset? what about those who have not written much on X?
@mohsinproduct Unfortunately, I don't use the X API directly, which makes scraping your reply history impossible for now.
That said, there's an alternative I've been thinking about: during onboarding, we could present you with different tweet scenarios and ask you to reply in your own words. That way, the tool could analyze your real tone of voice and syntax from scratch.
Definitely something worth exploring. Thanks for the feedback, this is gold! 🙏
That's actually really interesting. Aren't we limited to the topic shown in the image? Let's say, if I mostly engage with music-related threads, would that be possible?
Either way, it's really great work, well done!
@xencue Right now, yes, there are some limitations, but feedback like yours is super valuable to help shape and improve the product, so really appreciate you sharing this!
That said, the topic you select only influences the tone of voice, not the actual content. The extension understands the context of each tweet and generates replies based on that. So if you're engaging with music-related threads, it'll pick up on that and craft responses accordingly.
Give it a try and let me know how it goes!
@0x_alx Oh I see, that sounds nice, I'll definitely try it out! Thanks Alex!
@xencue 🫶🏼
50x impressions on X is impressive! The personalized voice profiles and image analysis features sound powerful. I'm curious about the AI behind this - which LLM powers the reply generation? Is it using Claude, GPT-4, or a custom fine-tuned model? How do you ensure the AI captures the user's authentic voice while maintaining engagement quality? 🐦
@new_user___01920256530b9f092c9e057
Thanks! Glad those features caught your eye 🙏
To answer your question, Warbl is powered by Gemini with heavily optimized prompting. No fine-tuning, just a lot of iteration on the prompts to get the tone and context right.
For capturing the user's authentic voice, it's all about layering: the voice profile settings define how you sound, while the AI analyzes the tweet content to understand what you're responding to. Keeping those two separate is key to avoiding generic-sounding replies.
Still refining it, but the results have been pretty solid so far!
Hello,
I like it, but you have an issue with your pro upgrade URL; it gives me a 500 error.
@viktorvillalobos Hey Victor ! Thank's for the feedback.
Is it still not working ? Can you DM me on X @indie_alx ?
aha, sign in google failed. Are you sure the product ready?
@tibelf fixing some last minute bugs ! almost 1000 replies generated and 4 paid users. th product works, need just some fix 😅
As an X user for marketing purposes, l can tell this tool solves a real issue but l do not get your business model in terms of monetization: you offer unlimited ai replies to premium users, so l can use this tool either 3-5 times a day or even 100+ times a day. The price l pay does not change. If a great number of your users use the tool for numerous times consistently each day, Then it will not be a profitable business. Additionally, there might be bot attacks increasing ai costs dramatically. Why did you choose to offer umlimited with fixed price
@eren_kiratli Really appreciate the thoughtful question, you're raising a valid concern and it shows you've got a sharp business eye 👀
Honestly, the unlimited model is a bet on simplicity and user experience. Most users won't hit 200+ replies a day, so the pricing is based on average usage patterns.
And for those power users who do engage heavily? They're likely getting real value and results from the tool, which means they'll stick around longer, that's a trade-off I'm willing to make.
As for bot attacks, there are rate limits and safeguards in place to prevent abuse.
That said, you're right that this model might evolve as usage data comes in. For now, I'd rather keep it simple and attractive for early adopters, then adjust if needed.
Thanks for the feedback, this is exactly the kind of thinking that helps refine the approach!
It is a great product as it is, but I would actually feel a lot comfortable using the tool if I knew the tool has scrapped all my replies in the past, and is using them as a reference to generate replies in my style and syntax!
@nabeeltak Love this idea and totally get where you're coming from! Having the tool learn from your actual past replies would be the dream.
Unfortunately, I don't use the X API directly, which makes scraping your reply history impossible for now.
That said, there's an alternative I've been thinking about: during onboarding, we could present you with different tweet scenarios and ask you to reply in your own words. That way, the tool could analyze your real tone of voice and syntax from scratch.
Definitely something worth exploring. Thanks for the feedback, this is gold! 🙏