XIST AIGC Studio — Breaking the Boundaries of Filmmaking with the Power of AI

XIST AIGC Studio：以AI之力，打破电影创作的边界

XIST AIGC Studio与COX SOUNDTRACK两家规模虽小却人才济济的公司——携手合作，将各自的优势融为一体，共同打造出一部以AI生成的短片，精彩地突破了传统制作模式的框架。

经过逾两百万次提示词的反复测试，以及对每个镜头数以千计的精心调整，只为追求最佳呈现效果，XIST AIGC Studio出品的《钳秦攻势》（The Qin Envelopment）已跻身年度最受瞩目的短片之列。影片将中国古代历史与动作武打、奇幻及科幻元素融为一体，使这部电影预告片大放异彩，成功入围新片场2025年度金雀奖（该奖项今年首次设立此类别）。

他们的创作理念与幕后故事究竟从何而来？TrueVisions Now特邀团队成员围坐畅谈，为有志于电影创作的人提供宝贵的参考与启发，让大家能够即刻将好创意付诸实践，不断精进自己的技艺。

请简要介绍 XIST AIGC Studio，包括目前团队成员数量，以及是什么原因或机会让团队走到一起的。

XIST AIGC Studio 目前团队规模在 5 人左右，正处于稳步扩充阶段，计划最终打造一支 10 人左右的精英团队。我们的核心成员大多是原影视广告行业的“全流程”或者“半流程’的「超级创作者」，基本每个人都会掌握多项技能，并且还在不断学习中，这样的特点让我们彼此减少大量的沟通成本，并且每个人在作品审美上都有较高的要求，正是这种价值观的一致性让我们走到一起，我们也希望未来有来自不同国家，对创作有同样价值观的人来一起创作。

《钳秦攻势》入选新片场金雀奖最佳AI创意短片，团队感觉如何？这次认可对团队及其未来工作方向有何意义？

《钳秦攻势》最初其实是我们基于 SD 2.0 模型进行的一次深度测试样片。能够入选新片场金雀奖，对团队来说是一个巨大的惊喜。这次认可不仅印证了我们建立的 AIGC 工业化工作流程是行之有效的，更坚定了我们未来的发展方向 —— 不盲目量产，而是聚焦在作品的口碑。

这部作品的主要灵感来源或核心想法是什么？

我们的核心想法是「顺应 AI 的特性进行原生创作」。在现阶段的 AI 创作中，精准控制历史服装或特定建筑依然存在技术门槛。因此，我们转变了思路：与其去死磕难以控制的元素，不如在编剧阶段就量身定制最适合 AI 表现的题材。比如混乱的时间线、多维空间的交错以及时代的重叠，这些恰恰是 AI 最擅长展现的视觉奇观。这部作品的灵感，正是源于这种「让故事适配技术」的逆向思维。

团队认为，这部作品如何提升视觉美感，或为娱乐行业带来新的维度，与过去主要由人类创作的作品相比有何不同？

过去，电影电视剧的创作话语权掌握在少数大型影视公司手里；而新的 AI 创作工具正在推动行业的「平民化」，未来必将涌现大量新兴工作室和独立创作者。在《钳秦攻势》中，我们最大的不同在于将传统的「电影视听语言」与前沿的「提示词工程」进行了深度融合。至于 AI 与人类的关系，我们始终认为 AI 只是加速工业流程的「工具」，提高了下限，但一部作品的灵魂、审美和最终高度，依然完全取决于「人」的决断。

团队在选择AIGC技术时有哪些标准？制作这部作品主要使用了哪些AI模型？团队大约花了多少时间来编写提示词（prompt）和反复调整，以获得满意的结果？团队如何处理许多AI模型中常见的局限性？

在 AIGC 这个飞速发展的未知领域，我们的标准就是「保持极致的敏锐」。只要是最新、最具潜力的方案，我们都会第一时间去尝试。这部作品核心使用的是春节期间刚刚发布的 SD 2.0 模型。整个制作周期非常紧凑，大约只花了 20 天。最终结果令我们非常满意，当然，过程中也遇到了算力紧张、模型火爆导致排队时间过长等局限性，这也是现阶段我们正在努力通过优化工作流来克服的技术痛点，好在我们很早就建立了AI Agent的工作习惯，会根据新模型的出现，不断更新迭代我们的工作流。

在制作这部作品的过程中，最大的挑战是什么？团队是如何克服它，让作品在美学和技术上都表现突出？

最大的挑战在于 SD 2.0 采用了全新的多模态方案，它的能力远超我们的预期，经常在生成过程中抛出意想不到的惊喜。但在影视制作中，我们需要极高的可控性。为了在这个充满随机性的模型中保证场景、人物和整体美术风格的「一致性」，团队投入了大量精力去适配该模型，开发并沉淀了一套我们审美专属的「提示词工程体系」，最终在美学表现和技术控制之间找到了平衡。

在AI创新方面，团队开发或应用了哪些技术和流程，为短片制作带来新鲜感？如果有特别成功的创新，请与我们分享。团队认为这些创新未来将如何影响电影行业的发展？

最成功的创新在于我们开发了专属的「全链路工作流」软件，覆盖了从创意、分镜到完整视频与音乐交付的每一个环节。这带来了一个根本性的范式转变：从传统的「为过程付费」转变为「为结果付费」。我们可以快速生成最终画面的雏形，先看到结果，再依据视觉反馈进行修改。这种迭代方式极大地加速了制作周期。我们预计，未来一部高标准的 AI 电影，完全可以把制作周期压缩到一个月以内。

与ALEX和COX SOUNDTRACK团队的合作是如何开始的？特别是ALEX在确定创作方向方面扮演了什么角色？团队在将音乐与AI生成的画面结合时，遇到了哪些挑战或发现了哪些优势？

ALEX 和 COX SOUNDTRACK 的合作源于对 AI 音乐的早期探索。在这个过程中，ALEX 更多地扮演了「AI Agent 开发者」的角色。我们在合作中确立了一种颠覆传统的创作逻辑：「音乐前置，画面填充」。因为在 AIGC 流程中画面的重置成本极低，所以我们决定让所有的视觉画面都去服务于音乐的情绪起伏。这种先定情绪基调、再用画面去共振的创作手法，能更好地向观众传达情感力量。

团队是否有计划将《钳秦攻势》的故事扩展成一部完整的AIGC长片？如果有，团队对这个项目的规划或设想是怎样的，以及预计何时能让观众看到？

我们确实有这个计划，并且目前已经进入了故事重写和剧本扩充的阶段。在规划上，我们采取了「小步快跑」的策略：首先会将《钳秦攻势》以「周更短剧」的形式推出，通过持续更新来积累观众反馈；到了合适的节点，再将其整合升级为一部完整的电影。由于目前团队也承接了一些 AI 影视的商业订单，长片计划会随着我们生产力的提升稳步推进，但大家期待的 AI 短剧版本已经正式提上日程。

根据团队的亲身经验，对于想成为AIGC创作者或开始制作自己作品的人，您有哪些建议？他们应该从什么开始做起？

在 AIGC 时代，所有人其实都站在同一起跑线上，摸索着一条全新的道路。每一个刚刚起步的创作者，都有可能成为某个全新流派的开创者。我们的建议是：传统影视的技术也许在被AI替代，但是如何达到高水准的思维路径依然保有强有力的竞争力。既然工具的门槛已被打破，只要你有想法，就一边实现，一边提高自己的思维和审美。行动是第一步。

XIST AIGC Studio — Breaking the Boundaries of Filmmaking with the Power of AI

XIST AIGC Studio and COX SOUNDTRACK — two small but exceptionally talented companies — have joined forces, combining their respective strengths to produce an AI-generated short film that beautifully shatters the conventions of traditional filmmaking.

With over two million prompt iterations tested and thousands of meticulous shot-by-shot refinements made in pursuit of the best possible outcome, The Qin Envelopment (钳秦攻势) by XIST AIGC Studio has emerged as one of the most closely watched short films of the year. By blending ancient Chinese history with action combat, fantasy, and sci-fi elements, this film trailer stands out so distinctly that it earned a nomination at the Xinpianchang 2025 Golden Sparrow Award — a category held for the very first time this year.

How did the concept behind this work come to life, and what went on behind the scenes? TrueVisions Now sits down with the team for an in-depth conversation — perfect for aspiring filmmakers looking to expand their knowledge, spark fresh ideas, and sharpen their craft right away.

Please briefly introduce XIST AIGC Studio, including the current number of team members and the reasons or opportunities that brought the team together.

XIST AIGC Studio currently has 5 members, aiming for an elite team of 10. Our core consists of multi-skilled "super creators" from the film and ad industries. Shared high aesthetic standards and low communication costs brought us together to push creative boundaries.

How does the team feel about The Qin’s Offensive being selected as one of the Best AI Creative Films at the Golden Sparrow Award? What does this recognition mean for the team and its future direction?

It started as an SD 2.0 test demo, so the nomination is a huge thrill. It validates our industrialized AIGC workflow and reinforces our core strategy: focusing on premium quality rather than blind mass production.

What was the main source of inspiration or key idea behind this work?

Our philosophy is "native AI creation." Instead of fighting AI's current limitations with strict historical accuracy, we write stories tailored to its strengths—such as chaotic timelines, multidimensional spaces, and overlapping eras.

In the team’s view, how does this work improve visual aesthetics or bring new dimensions to the entertainment industry compared to most past works created mainly by humans?

AI democratizes the industry, shifting power from major studios to independent creators. We stand out by fusing traditional cinematic language with prompt engineering. AI raises the baseline, but the soul and ultimate height of a work still rely entirely on human aesthetic judgment.

What criteria does the team use when choosing AIGC technologies? Which main AI models were used to produce this work? Approximately how much time did the team spend on creating prompts and doing iterations to achieve satisfying results? How did the team handle the common limitations found in many AI models?

We always embrace the newest tools. We used the recently released SD 2.0 for a tight 20-day production. To overcome common limitations like compute constraints and queue times, we rely on AI Agents to continuously update and optimize our workflow.

From the experience of producing this work, what was the biggest challenge? How did the team overcome it to make the work stand out in both aesthetics and technical quality?

The main challenge was taming SD 2.0’s randomness. To ensure visual consistency across scenes and characters for film standards, we developed a proprietary "prompt engineering system" to perfectly balance artistic expression with technical control.

In terms of AI innovation, what techniques or processes did the team develop or apply that bring freshness to short film production? If there is any particularly successful innovation, please share it with us. How does the team think these innovations will influence the future development of the film industry?

We developed an exclusive "full-pipeline workflow" software. It shifts the paradigm from "paying for the process" to "paying for the outcome." By generating visual prototypes for immediate feedback, we can compress a feature film's production cycle to under a month.

How did the collaboration with ALEX (柿子_ALEX) and the COX SOUNDTRACK team begin? In particular, what was ALEX’s role in shaping the creative direction? What challenges or strengths did the team find when combining music with AI-generated visuals?

ALEX acts as our AI Agent Developer. Together, we adopted a "music first, visuals second" approach. Since regenerating AIGC visuals is incredibly cheap, we let the imagery entirely serve the music's emotional arc, maximizing the overall impact.

Does the team have plans to expand the story of The Qin’s Offensive into a full-length AIGC film? If so, how has the team planned or envisioned this project, and when do you expect it could be presented to audiences?

Yes. We are currently expanding the script using an agile strategy: we will first release it as a weekly short series to gather audience feedback, and eventually upgrade it into a full feature film.

Based on your direct experience, what advice do you have for people interested in becoming AIGC creators or starting to make their own works? What should they begin with first?

Everyone is at the same starting line. While AI lowers technical barriers, high-level storytelling and aesthetics remain your core competitive edge. Don't wait—just start creating, and refine your taste along the way.