Ivo R.
Title: Misleading Features, Poor Support – Avoid if You're Expecting Professional Results
I signed up to Kaiber hoping to design a band logo using their AI video and image tools, but the experience was frustrating from the start. Despite carefully crafted prompts and uploading my own sketches, the results were distorted and unusable. It felt like intentional sabotage forcing you to buy more credit. The platform gives you no upfront warning that certain input types—like hand-drawn logos—won’t work, nor does it guide you toward the right tools before you waste your credits. When I contacted support to request a refund for my unused credits, I was met with condescending, AI-generated replies that blamed me for not using the platform "correctly," instead of acknowledging the lack of clear onboarding or proper tool descriptions. Even after escalating, I was refused a refund and told it was essentially my fault for assuming the tools would work as expected. Unless you're prepared to waste time and money experimenting blindly, I would steer clear. Kaiber may have potential, but the current setup and customer support make it a poor investment.
Response from Kaiber SuperstudioJun 24, 2025
The features are not misleading. The issue stems from how they were used. Specifically, you attempted to create a Kling video using a sketch on lined paper, which you were previously advised not to do. We provided clear guidance on how to prepare images correctly for that feature, but that advice was rejected aggressively. In addition, the prompts you submitted were not suitable for any AI video generation—they lacked structure, clarity, and essential descriptive elements. We also took the time to advise you on how to craft effective prompts that align with the system’s capabilities. The reality is this: unless you're open to learning from these experiences, it will be difficult to improve your results or grow as an AI creator. Tools like Kling are powerful, but they require accurate input and a basic understanding of how they work. We do not issue refunds for content generated using incorrect or unsuitable input. That said, we still offered you a goodwill credit to retry using the proper approach and guidance provided—but you declined that support. We are not responsible for the users prompt but will guide the user to correctly structure their prompt. We're always here to help users get the most out of the platform, but we also expect a willingness to engage constructively and follow the tools' intended use. For context i even broke down the issues you were facing 1. The Prompts: Your prompts are too abstract and poetic, using design language that lacks clear visual direction. Terms like “timeless elegance” or “surreal yet refined” sound appealing but don’t tell the AI what to generate. For successful results, prompts need to be visually specific, focusing on elements like shape, composition, style, lighting, and background. Without this clarity, the AI struggles to produce anything meaningful. 2. The Image (Lab 2): You uploaded a sketch on lined paper into Image Lab 2, which isn’t suited for refining rough drawings. The notebook lines interfere with the model’s interpretation, and the sketch lacks definition for proper stylization. For this type of input, Image Lab 1 with the Stencil Upload feature is the correct tool. Using a vague prompt with an unclear sketch gives the AI poor data to work with, which leads to disappointing results. 3. Kling Video You’ve also tried animating the same lined-paper sketch in Kling, which isn’t going to produce the effect you're expecting. Kling does not restyle or auto complete sketches—it simply animates the image exactly as it is. If the source image is rough, incomplete, or includes distracting elements like paper lines, that’s exactly what will be animated. The output will feel broken or unfinished because the input is. Hopefully you will take onboard the information given and have a better generative experience in ai. or at least this post serves a purpose at helping advise others. thank you