happyhorse nsfw on Anyken: compare the model examples with your intended input and output. Review the capabilities documented below, then use the controls actually offered in the workspace. Available settings and credits are shown before submission.
Read source
HappyHorse 1.0 is an AI video generation model developed by Alibaba Token Hub's innovation business unit, available on fal as an official API partner. Happyhorse nsfw supports text-to-video, image-to-video, video editing, and reference-to-video endpoints, generating 1080p output with synchronized native audio. This section covers the model's architecture, team background, and what the documentation says about content policies.
HappyHorse 1.0 uses a unified 40-layer self-attention Transformer with 15 billion parameters, generating video and audio jointly in a single forward pass without separate audio post-processing. Happyhorse nsfw supports four input modalities — images, videos, audio, and text — for richer expression and more controllable generation. Users can upload up to 9 images (jpeg/png/webp/bmp/tiff/gif, max 30MB each), up to 3 videos (mp4/mov, total duration 2-15s, max 50MB each), and up to 3 audio files (mp3/wav, max 15s duration, max 15MB each). The mixed input limit is 12 files total. The model generates 1080p video with synchronized native audio and supports multilingual lip-sync across seven languages: Mandarin, Cantonese, English, Japanese, Korean, German, and French. The documentation does not describe any built-in content filtering or NSFW detection mechanism.
HappyHorse 1.0 was developed by the Future Life Lab team at Taotian Group (Alibaba), led by Zhang Di, a veteran AI engineer with 15+ years in the field. Zhang Di served as Director at Alibaba Group from 2010–2022, then joined Kuaishou as Vice President where he was the technical architect of Kling AI. He rejoined Alibaba in late 2025 to lead the Taotian Future Life Lab under Alibaba Token Hub (ATH), and within months delivered HappyHorse 1.0. Happyhorse nsfw was submitted pseudonymously to the Artificial Analysis leaderboard on April 7, 2026, immediately ranking #1 in both Text-to-Video and Image-to-Video (no audio) categories based on blind human preference votes. The model is closed source and not licensable. The documentation does not establish any content moderation policies or NSFW content restrictions.
HappyHorse 1.0 offers four main endpoints via the fal API: text-to-video, image-to-video, video editing, and reference-to-video. Happyhorse nsfw generates 1080p video with synchronized native audio and supports advanced features like video editing through natural language instructions. This section details the documented capabilities and access paths, noting that the sources do not specify any content policy or NSFW filtering.
The HappyHorse 1.0 API provides four endpoints: text-to-video (alibaba/happy-horse/text-to-video) generates 1080p video with synchronized native audio from a text prompt, supporting aspect ratios 16:9, 9:16, 1:1, 4:3, 3:4 and durations 3-15 seconds. Image-to-video (alibaba/happy-horse/image-to-video) generates video with synchronized native audio and multilingual lip-sync from text prompts or images. Video editing (alibaba/happy-horse/video-edit) supports local or global editing of video elements using up to 5 reference images through natural language instructions. Reference-to-video (alibaba/happy-horse/reference-to-video) generates 1080p video with synchronized native audio from a text prompt and references. Happyhorse nsfw also supports stylized transform and lipsync features. The documentation does not describe any content moderation or NSFW filtering for these endpoints.
HappyHorse 1.0 is available on fal via both the playground (no API key needed) and the API. The playground allows users to generate videos instantly without setup. For developers, the API can be integrated using Python and JavaScript SDKs, plus a REST API for any language. Pricing is $0.14/second of generated video at 720p and $0.28/second at 1080p, with pay-per-second billing and no minimums or subscriptions. Enterprise pricing is available through sales contact. Additionally, Alibaba Cloud offers a limited-time 40% discount on Pay-As-You-Go rates for HappyHorse models, with 720p starting at $0.084/sec and 1080p at $0.108/sec. A free trial of 10-second video generation is available through Alibaba Cloud Model Studio. The pricing documentation does not mention any content-based restrictions or NSFW content policies.
HappyHorse 1.0 is available on fal as an official API partner and through Alibaba Cloud Model Studio. Happyhorse nsfw supports text-to-video, image-to-video, video editing, and reference-to-video endpoints. This section covers the current availability status, platform access methods, and what the documentation says about content policies.
HappyHorse 1.0 is available now on fal as an official API partner. All four endpoints — text-to-video, image-to-video, video editing, and reference-to-video — are live in the playground and via the API. Happyhorse nsfw is also available through Alibaba Cloud Model Studio, which offers a free trial of 10-second video generation. The Alibaba Cloud campaign runs from launch until July 6, 2026, offering a 40% discount on Pay-As-You-Go rates for HappyHorse models. The model is closed source and not licensable. The documentation does not specify any content moderation policies or NSFW content restrictions for any of these access paths.
The fal platform provides a client API that handles request submission, status updates, and result retrieval. Integration is available via Python and JavaScript SDKs, with code examples provided for text-to-video and image-to-video endpoints. The API uses fal.subscribe() to submit requests with a prompt and parameters, automatically handling the async queue and returning the final video URL when generation is complete. Alibaba Cloud Model Studio offers an integrated workflow from AI generation to fine editing, with preset prompt templates for quick access. The documentation includes API documentation, prompting guides, and sample outputs from the Artificial Analysis Video Arena. None of the platform documentation describes content moderation policies or NSFW filtering mechanisms.
HappyHorse 1.0 offers configurable parameters for video generation, including resolution, aspect ratio, duration, and input modalities. Happyhorse nsfw supports multiple input types and provides flexibility in output quality. This section details the documented technical configurations and parameters.
HappyHorse 1.0 supports four input modalities: images (up to 9, formats jpeg/png/webp/bmp/tiff/gif, max 30MB each), videos (up to 3, total duration 2-15s, formats mp4/mov, max 50MB each, resolution 409600-927408), audio (up to 3 files, max 15s duration, formats mp3/wav, max 15MB each), and text (natural language prompts, generation length freely adjustable from 4-15 seconds). The mixed input limit is 12 files total. Output parameters include resolution options (720p and 1080p), aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4), and duration (3-15 seconds for text-to-video and reference-to-video endpoints). The model generates video and audio jointly in a single forward pass, producing synchronized audiovisual output without separate audio post-processing.
The HappyHorse 1.0 team claims approximately 38-second generation time for 1080p output on a single NVIDIA H100 GPU, and roughly 2 seconds for a 5-second clip at 256p. The model uses a unified 40-layer self-attention Transformer with 15 billion parameters and no cross-attention modules. The model holds the #1 Elo rating on the Artificial Analysis Video Arena in both Text-to-Video (no audio) with Elo 1333 and Image-to-Video (no audio) with Elo 1392, based on blind human preference votes. For Text-to-Video (with audio) the model ranks #2 with Elo 1205, and for Image-to-Video (with audio) it ranks #2 with Elo 1161. These scores reflect early vote counts from April 2026 and may shift as more votes accumulate. The documentation does not describe any content-based performance variations or filtering mechanisms.
This section provides practical guidance for users evaluating HappyHorse 1.0 for various use cases. The documentation establishes technical capabilities and access paths but does not specify content policies. Users should consult current service terms for definitive policy information.
The official HappyHorse 1.0 documentation, including the fal platform page, the happy-horse.ai documentation, and the Alibaba Cloud campaign page, does not establish an explicit adult-content policy or built-in NSFW filtering mechanism. The documentation describes technical capabilities, pricing, access paths, and performance specifications but does not mention content moderation, content filtering, or NSFW detection. Users who need to understand content policies for specific use cases should consult the current service terms of the hosting platforms (fal and Alibaba Cloud) directly. The documentation does not provide information about account suspensions, regional content restrictions, or credit policies related to content types.
To get started with HappyHorse 1.0, users can access the fal playground without an API key to generate videos instantly. For API integration, the process involves installing the fal client (npm install @fal-ai/client for JavaScript or pip install @fal-ai/client for Python), creating an account on fal, obtaining an API key from the developer dashboard, and submitting requests using fal.subscribe(). The model supports multilingual lip-sync across seven languages: Mandarin, Cantonese, English, Japanese, Korean, German, and French. Users can test these capabilities in the fal playground. The documentation includes a prompting guide with tips for camera moves, lighting, and multi-beat shots. For enterprise needs, users can contact sales for pricing and support. The documentation does not provide guidance on content policy compliance or NSFW content handling.
Frequently asked questions about HappyHorse 1.0, based on the official documentation.
The official HappyHorse 1.0 documentation does not establish an explicit adult-content policy or built-in NSFW filtering mechanism. The cited sources describe the model's technical capabilities, pricing, access paths, and performance specifications but do not mention content moderation, content filtering, or NSFW detection. Users who need to understand content policies for specific use cases should consult the current service terms of the hosting platforms (fal and Alibaba Cloud) directly. The model supports text-to-video, image-to-video, video editing, and reference-to-video endpoints, generating 1080p video with synchronized native audio. The documentation does not provide information about account suspensions, regional content restrictions, or credit policies related to content types.
HappyHorse 1.0 holds the #1 Elo rating on the Artificial Analysis Video Arena in both Text-to-Video (no audio) with Elo 1333 and Image-to-Video (no audio) with Elo 1392, based on blind human preference votes where real users compare outputs without knowing which model produced them. For Text-to-Video (with audio) the model ranks #2 with Elo 1205, and for Image-to-Video (with audio) it ranks #2 with Elo 1161. These scores reflect early vote counts from April 2026 and may shift as more votes accumulate. The rankings are based on the Elo rating system where users compare two unlabeled clips and pick the better one, with the winning model gaining points and the loser losing some.
HappyHorse 1.0 pricing on fal is $0.14/second of generated video at 720p and $0.28/second at 1080p, with pay-per-second billing and no minimums or subscriptions. Enterprise pricing is available through sales contact. Additionally, Alibaba Cloud offers a limited-time 40% discount on Pay-As-You-Go rates for HappyHorse models, with 720p starting at $0.084/sec and 1080p at $0.108/sec. A free trial of 10-second video generation is available through Alibaba Cloud Model Studio. The Alibaba Cloud campaign runs from launch until July 6, 2026, and is open to users registered on the Alibaba Cloud official website. The discount does not apply to Prepaid subscriptions or AI General Savings Plans.
