AI Card Bao HERMES.AICARDBAO.COM

Industry News

Kimi came up with a game development plan: For planners who don’t know how to code, it’s possible to create a playable prototype within 1 to 2 days.

What Kimi released this time isn’t actually a new model, but rather an industry-specific solution: K3.1 Pro combined with Kimi Code and Kimi Work, along with tool plugins for Unity, Unreal Engine, and Blender – all of which enable direct integration of these models into existing projects.

What it focuses on is the period before a project is officially launched. Whether a particular gameplay concept is actually interesting can only be determined after it has been actually tested and played, but creating a prototype requires both resources and time. The official cited Unity’s 2026 report: 52% of developers prefer to start with small-scale projects to reduce risks, while 67% strive to keep the prototype development phase within three months. According to Kimi, a planning or creative team without any programming experience can create a playable prototype in just 1 to 2 days.

The officials provided two examples. The first example involved adding rule cards that could be freely combined into the three-match elimination gameplay. First, they analyzed how these rule combinations affected scoring and strategic play, and then broke down these effects into aspects related to interaction, numerical values, and interface elements. The second example involved using reference images and videos to create a 2.5D action game demo using Godot. This demo needed to include features such as rain effects, water on the ground, as well as actions like swinging a sword, dodging, and blocking attacks. They also ran the game in actual operation to examine the rendered visuals and trial-play recordings, making necessary adjustments whenever issues were identified.

The latter half of the tasks was also included: organizing player feedback, preparing materials for acquiring new customers, implementing localization efforts, and conducting preliminary screening of user-generated content.

These are all just manufacturer demonstrations. There is still a long way to go between making a prototype that is “ playable” and one that is “so much fun that it’s worth investing resources in developing further.” The reliability of making code changes within large-scale projects also needs to be verified separately. The official recommendation, however, is quite practical: start with tasks that are easy to check and verify.

Related