AIGCLISTAIGCLIST
Viggle AI
AI 工具评分卡

Viggle AI

Viggle AI:物理级动作迁移,一张图让任何角色动起来

免费增值AI 视频生成器viggle.ai
访问
发布于 2026年7月6日

基准评分

Viggle AI 在 Agent 就绪度与 AI 可见性上的得分 AI 就绪度和 GEO Score 是 VibeLaunch 在提交后生成的平台评估。

由 AIGC List 基准评分提供支持

决策摘要

Viggle AI

适合

  • API-first design enables straightforward developer integration with no specialist tools required.
  • Large library of pre-built motion templates reduces the need for custom animation sourcing.
  • PINOC eliminates the need for expensive motion capture hardware for skeletal animation extraction.

注意

  • All generation quality and speed claims are vendor assertions without independent third-party verification.
  • Pricing beyond the free trial tier is not publicly documented, creating uncertainty for budget planning.
  • The sub-20-second generation claim lacks independent throughput or latency benchmark data.

概述

Viggle AI 首页展示角色动画示例

2024年3月,一段15秒的视频在TikTok上病毒式传播:华金·菲尼克斯版的小丑走上了Summer Smash音乐节舞台,取代了说唱歌手Lil Yachty的出场。步态精准。夹克的摆动分量刚好。角色逐帧锁定在动作上。这不是换脸,而是从一张静态图和一段参考视频生成的全身体动作迁移,出自当时还鲜为人知的多伦多初创公司 Viggle AI。公开上线五个月内,平台Discord社区成员达到400万,据TechCrunch报道

Viggle 的核心判断是:对大量创作者而言,精准比炫目更重要。一段舞蹈短视频、一个 meme 模板、一个直播虚拟形象:这些场景要求角色精确复现动作,而不是大致像那么回事。RunwayKling 3.0 等通用视频生成工具可以从文字提示词产出电影级画面,但它们不解决"让特定角色做出特定动作,且关节精度到位"的问题。

JST-1:物理引擎而非像素插值

Viggle 的底层模型 JST-1,公司称之为视频-3D基础模型。与依赖像素模式插帧的方法不同,JST-1 从参考视频中重建3D骨骼,再将其映射到目标角色图像上,期间遵循关节约束、足部接触和布料物理。公司称其为首个"真正理解物理"的同类模型,,ChatForest的评测指出,这个说法比大多数AI营销用语更经得起检验:输出结果展示的是正确的重量分布,而不仅仅是视觉上过得去。

技术背景值得关注。创始人兼CEO Hang Chu 是计算机视觉研究员,曾在Autodesk、Meta、NVIDIA和Google任职,2022年在多伦多创办 Viggle。他的学术背景包括在多伦多大学师从 Raquel Urtasun(后创办自动驾驶公司Waabi)和 Sanja Fidler(现任职NVIDIA)的博士研究。截至2026年初,团队规模约31至36人,base在多伦多。

能做什么

Viggle 的功能集已远超最初的 meme 生成器。核心模式包括:

  • Mix:上传一张角色图和一段动作视频,Viggle 将动作迁移到角色上。那个小丑视频就是 Mix 模式的作品。
  • Multi:同一场景中动画化多个角色,每个角色可接入独立的动作源。
  • Move:用文字提示词而非参考视频来引导角色动作。
  • Real-Time Swap:直播中的面部和身体替换。免费版每天5分钟,Max 套餐无限制。

2025年,Viggle 推出了 PINOC,一个独立工具,可以从任意视频画面中提取骨骼动画数据,导出为 FBX 或 GLB 文件供3D软件使用。同期还发布了 Fight Anyone 3D,一款AI驱动的格斗游戏,用同一生成模型实时渲染自定义角色的对战场景。

平台提供 API,文档见 docs.viggle.ai,供开发者将动作迁移集成到自有应用中。

公司与融资

Viggle 于2024年8月完成1900万美元A轮融资,由 Andreessen Horowitz(a16z)领投,Two Small Fish Ventures 参投,据BusinessWire。总融资额约2700万加元。公司宣称截至2026年拥有超过4000万用户,涵盖网页端和移动端。Discord 仍是主要社区阵地,公司在关于页面上强调其 Discord 排名。

价格(截至2026-07-09)

套餐 月费 点数 并发生成 水印 实时换脸
Free $0 无(5个视频/天) 1 5分钟/天
Pro $9.99(年付$7.99) 80 4 30分钟/天
Live $19.99(年付$15.99) 200 6 400分钟/天
Max $79.99(年付$63.99) 800 10 无限制

点数按生成消耗,不同模型和功能消耗点数不同,据Viggle定价FAQ。月付点数在账单日过期,购买的点数不过期但需保持有效订阅才能使用。

定位

Viggle 占据了一个介于多个大品类之间的利基。它不是文字生成电影级画面的工具(那是 Runway、Kling 的领地),也不是数字人播报平台(HeyGenSynthesia)。在精神上最接近的是 Pixverse AI 等动作控制工具,但 Viggle 更侧重物理保真度和角色一致性,而非场景多样性。

这种专注有其代价。输出分辨率低于专用电影级生成器。免费套餐在服务器繁忙时生成速度会受影响。复杂的多角色遮挡场景可能出现可见伪影。工具在源动作干净、角色居中且光线充足的条件下表现最佳,这些条件更利于棚拍式内容而非随手拍素材。

竞品对比

Runway Gen-4 和 Kling 3.0 产出更高分辨率的电影级画面,支持文字生成场景,这是 Viggle 不参与竞争的领域。HeyGenSynthesia 聚焦企业视频的数字人播报,属于不同使用场景。在动作专用工具中,Pixverse AI 提供角色动画但偏向风格化输出,而非物理精确的动作迁移。

Viggle 的优势在于骨骼追踪的精度。当角色需要精确复现参考视频中的舞蹈动作、步态循环或手势时,Viggle 能提供逐帧级别的保真度,这是通用生成器不优先考虑的指标。代价是更窄的能力范围:无场景生成,无文字转视频,输出分辨率较低。对只需要那段动作的创作者而言,这个取舍是成立的。

Viggle 的 Discord 服务器在上线五个月内达到400万成员,至今仍是核心分发渠道。用户在社区频道分享模板、动作片段和二次创作。公司定期举办主题挑战赛,在 TikTok 和 X 上产出病毒式内容,形成"病毒输出驱动新注册"的增长闭环。

局限

Viggle 不生成文字描述的场景,不产出高分辨率电影级画面。免费版带水印,且一次只能生成一个视频。Real-Time Swap 需要稳定网络,对正面、均匀光照的主体效果最好。公司宣称的4000万用户数据为自报数据,包含免费注册用户,活跃用户数缺乏独立验证。

评价 (0)

0 条评分

还没有评价。成为第一个评价的人!

评分构成

编辑评分由哪些维度构成,每项附判断依据。 AI 就绪度和 GEO Score 是 VibeLaunch 在提交后生成的平台评估。

Information quality

Developer documentation is clear and functional but every capability claim originates from the vendor with no independent review or third-party validation available.

4.0
建议核验

All passages across homepage and developer docs are first-party vendor content. No third-party benchmarks, user reviews, or independent quality assessments are present in the source packet.

Ease of use

Single REST call design with Bearer token auth and no specialist tools required suggests low integration friction. Documentation covers the core workflow clearly in both English and Chinese.

7.2
依赖场景

Documentation describes a three-step workflow — upload photo, choose motion, receive MP4 — with Bearer token authentication. Dual-language developer docs available.

Feature depth

Core character swap plus template library, custom video support, and PINOC skeletal extraction represent meaningful feature breadth. Template count claimed in thousands but unverified.

6.2
建议核验

Features span consumer use (selfie-to-meme templates) and professional use (FBX/GLB export from PINOC). No detail on template curation, quality tiers, or resolution options.

Workflow fit

API-first design suits developer integration well. PINOC addresses a distinct professional workflow. The dual consumer-pro positioning may dilute focus but broadens addressable use cases.

6.5
建议核验

Documentation explicitly targets developers integrating into apps while homepage messaging addresses casual creators. PINOC targets 3D artists and animators with FBX/GLB export.

Reliability

No uptime data, SLA documentation, error handling detail, or independent latency benchmarks available. The sub-20-second claim is unverified and production behavior under load is unknown.

3.0
建议核验

Vendor claims under-20-second generation but provides no throughput limits, error codes, retry logic, or availability guarantees in the available documentation passages.

Value

Free trial is a positive signal for evaluation, but the complete absence of paid tier pricing, rate limits, or usage quotas in public documentation prevents meaningful value assessment.

3.5
建议核验

Homepage promotes TRY FOR FREE. No pricing page, tier breakdown, or usage quota information is present in the source packet beyond the free trial reference.

评分反映可查证的产品资料,不代表实际使用效果保证。

Agent 就绪度

评估 Agent 能否通过产品的官方信息理解产品,并重建一条有文档依据的工作流程。

Automated agent-readiness assessment of https://viggle.ai: 9 of 22 checks verified across 5 fetched pages. No substantial machine interface is documented — agents can understand and cite the product but not operate it. Absent: agent_tooling_artifacts, request_examples, version_information, changelog, cli, cli_non_interactive.

就绪度维度

评估维度得分
文档质量85
执行结果可验证性33
机器接口25
项目定位清晰度75
资源可发现性100
工作流完整度88

对 Agent 有帮助的部分

  • docs: verified during this run
  • llms txt: verified during this run
  • sitemap: verified during this run
  • quickstart: verified during this run
  • api reference: verified during this run
  • authentication: verified during this run

Agent 受阻的部分

  • No agent instruction files, code-distribution commands, or named slash-command skills found across fetched pages.
  • No request examples signal matched across 5 fetched pages.
  • No version information signal matched across 5 fetched pages.
  • No changelog signal matched across 5 fetched pages.
  • No cli signal matched across 5 fetched pages.
  • No sdk signal matched across 5 fetched pages.

证据核查

关于该工具的公开声明,每条均标注核验状态与引用来源。

viggle-ai12
viggle.ai已验证核验于 2026年7月13日

Viggle AI allows users to remix anyone into viral memes by transferring motion from video footage onto a character image.

PINOC extracts skeletal animation from any video footage without requiring a mocap suit or studio, supports 3D preview, and exports as FBX or GLB.

Fight Anyone 3D is a fighting game powered by Viggle's 3D generation model, supporting character customization, AI-generated 3D motions and scenes, and PvE, PvP, and friend battles.

Viggle AI and Fight Anyone 3D are available to try for free.

Viggle's Creator platform includes MIX multi-track motion control for coordinating complex character animations.

Viggle supports real-time character swapping within generated content.

Viggle's platform includes image generation and video generation capabilities alongside its core animation tools.

Viggle's product suite includes the Viggle creator app, Viggle Game Portal, and PINOC animation tool.

Viggle maintains a community with a Discord channel and regularly showcases user-generated content under Community Highlights.

Viggle V4 is the current motion control system iteration with a dedicated guide covering motion control capabilities.

Viggle partnered with AWA Studios to create an interactive fan experience for Ronda Rousey's graphic novel.

Viggle is operated by Viggle Inc. with copyright through 2026 and maintains Terms of Service, Community Rules, and a Privacy Policy.

https://viggle.ai
developers9
viggle.ai已验证核验于 2026年8月30日

Viggle AI provides a character-to-video swap API that places an extracted character into any video via a single REST call.

The API extracts a subject from an uploaded reference photo and prepares it for animation.

Users can choose from thousands of pre-built motion templates for character animation.

Users may upload their own driving video instead of using a pre-built motion template.

The API returns an animated MP4 in under 20 seconds after combining character and motion source.

API authentication uses a Bearer token generated from the developer portal at portal.viggle.ai.

The product supports letting users upload a selfie to appear in trending dance or meme templates.

The API requires no specialist tools or software beyond standard REST client capabilities.

A documentation surface is reachable at https://viggle.ai/developers.

https://viggle.ai/developers
zh-CN/developers8
viggle.ai厂商声明核验于 2026年7月16日

Viggle AI provides a character-to-video swap API that places an extracted character into any video via a single REST call.

The API extracts a subject from an uploaded reference photo and prepares it for animation.

Users can choose from thousands of pre-built motion templates for character animation.

Users may upload their own driving video instead of using a pre-built motion template.

The API returns an animated MP4 in under 20 seconds after combining character and motion source.

API authentication uses a Bearer token generated from the developer portal at portal.viggle.ai.

The product supports letting users upload a selfie to appear in trending dance or meme templates.

The API requires no specialist tools or software beyond standard REST client capabilities.

https://viggle.ai/zh-CN/developers
viggle.ai4
viggle.ai已验证核验于 2026年8月30日

PINOC extracts skeletal animation data from video footage and exports results as FBX or GLB files with 3D preview.

Viggle AI offers a free trial tier, promoted with a TRY FOR FREE call-to-action on the homepage.

Viggle AI positions itself for both serious technical applications and casual entertainment use.

The entry page was fetched and analyzed for machine-interface signals (title, headings, developer links, keyword probes).

https://viggle.ai/
Quick Start: Video Remix - Viggle API2
viggle.ai已验证核验于 2026年8月30日

A quick-start / agent-skills documentation page is reachable at https://docs.viggle.ai/v1/guides/quickstart-mix.

Agent-native positioning with a concrete operational path: "The documentation provides a concrete quickstart with API key setup and runnable code examples, which is an operational path for agents.".

https://docs.viggle.ai/v1/guides/quickstart-mix
https://viggle.ai/llms.txt1
viggle.ai已验证核验于 2026年8月30日

llms.txt is published at the site root and readable.

https://viggle.ai/llms.txt
https://viggle.ai/sitemap.xml1
viggle.ai已验证核验于 2026年8月30日

sitemap.xml is reachable and lists site pages.

https://viggle.ai/sitemap.xml
https://viggle.ai/_build/assets/entry-client-rE_pxMmV.css1
viggle.ai已验证核验于 2026年8月30日

A quick-start / agent-skills documentation page is reachable at https://viggle.ai/_build/assets/entry-client-rE_pxMmV.css.

https://viggle.ai/_build/assets/entry-client-rE_pxMmV.css
https://viggle.ai/_build/assets/entry-client-DCHWNLx6.js1
viggle.ai已验证核验于 2026年8月30日

A quick-start / agent-skills documentation page is reachable at https://viggle.ai/_build/assets/entry-client-DCHWNLx6.js.

https://viggle.ai/_build/assets/entry-client-DCHWNLx6.js

决策核对台

在依赖该产品或访问官网前,最值得先确认的问题。

Viggle AI 是一个角色动画与动作迁移平台。用户上传一张任意风格的角色图片和一段参考动作视频,Viggle 的 JST-1 模型通过物理感知的3D骨骼追踪,生成该角色做出参考动作的短视频。

是的,Viggle 提供免费套餐:每天5次视频生成、单次生成一个、输出带水印。付费套餐从 $9.99/月(Pro)起步,去除水印并增加点数和更快的生成队列。

Viggle 的 JST-1 模型从参考视频中重建3D骨骼结构,再将其映射到目标角色图像上,同时遵守关节约束、足部接触和布料物理。这与基于像素插值的方法不同,后者可能输出看起来合理但物理上错误的动作。

Viggle 的 Move 模式支持用文字引导动作,但平台的核心设计是参考视频到角色图像的动作迁移。它不是通用文字生成视频工具。如果需要电影级场景生成,Runway 或 Kling 更合适。

PINOC 是 Viggle 推出的独立工具,无需动捕服或专业设备即可从任意视频中提取骨骼动画数据。用户可3D预览动画并导出为 FBX 或 GLB 文件,用于 Blender、Unreal Engine 等3D软件。

请在官网核验

继续探索

相近任务的不同路径

这些工具以带有明确编辑理由的替代关系关联到当前产品。

01Pixelverse AI

Pixelverse AI

Offers similar character-to-video animation capabilities and may appeal to users comparing API-driven animation tools.

查看档案
02Aitubo

Aitubo

Provides AI-driven video creation in the same category; may suit users with different workflow preferences or output requirements.

查看档案
03Motion Control AI

Motion Control AI

Overlaps partially with Viggle AI's template-driven motion approach and may appeal to users who prioritize motion control over character swapping.

查看档案
查看 Viggle AI 的全部替代工具