Alibaba Cloud

Happy HorseAlibaba Cloud

Unlike fragmented toolchains that require multiple models stitched together, Happy Horse operates as a cohesive system. It enables consistent visual quality, temporal coherence, and style control across all generation modes, making it particularly valuable for production environments where reliability matters as much as creativity.

Happy Horse 1.1Alibaba Cloud

Happy Horse 1.1 is Alibaba Cloud's video generation family with highly realistic, fluid, and detailed motion. It automatically routes to text-to-video, image-to-video, or reference-to-video based on the inputs you provide.

Qwen 2 1.5B InstructAlibaba Cloud

Qwen2 1.5B Instruct: Efficient, multilingual language model for diverse NLP tasks.

Qwen 2 Instruct (72B)Alibaba Cloud

State-of-the-art AI model improving customer engagements through sophisticated chat functionalities.

Qwen 2.5 7B Instruct TurboAlibaba Cloud

Qwen 2.5 7B Instruct Turbo excels in coding and instruction following.

Qwen 3 235B A22BAlibaba Cloud

Advanced multilingual AI with 128K-token context, excelling in coding, reasoning, and enterprise applications.

Qwen 3.6 27BAlibaba Cloud

Qwen3.6-27B is the first fully dense open-weight model in the Qwen3.6 series and it just beat a 397-billion-parameter model on agentic coding benchmarks.

Qwen 3.7 MaxAlibaba Cloud

Qwen3.7-Max is Alibaba’s flagship reasoning and agentic LLM optimized for coding, productivity, long-context understanding and autonomous workflows. Text input and text output only. Supports function calling, explicit context cache and advanced reasoning capabilities.

Qwen 3.8Alibaba Cloud

Qwen 3.8 Max is Alibaba's flagship multimodal reasoning and agentic LLM. It understands video, images and text and returns text, optimized for coding, long-context tasks and autonomous workflows. Supports function calling, streaming and explicit thinking (reasoning) mode.

Qwen 3.8 27BAlibaba Cloud

Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It accepts text, image and video inputs and is suited for coding, professional workflows, research, multimodal interaction and long-running agent tasks, with flexible thinking that can be tuned per request.

Qwen ImageAlibaba Cloud

Qwen Image is a powerful multimodal AI model specialized in generating high-quality images from textual descriptions, built with 20 billion parameters.

Qwen Image 2.0Alibaba Cloud

Qwen Image 2.0 turns plain text into 2K photoreal visuals, polished posters, infographics, and slide-ready graphics with text that's actually readable.

Qwen Image 2.0 ProAlibaba Cloud

Unlike conventional AI image tools that focus mainly on aesthetics, Qwen Image 2.0 Pro emphasizes usability — clear typography, coherent layout systems, and predictable composition that works in real-world marketing and design workflows.

Qwen Image 3Alibaba Cloud

Qwen Image 3 is Alibaba's third-generation image generation model supporting up to 4,500-token prompts — 4.5x more than the previous version. Generates complex layouts including infographics, storyboards, UI mockups, and knowledge diagrams in one shot. Renders text as small as 10 pixels with precision, natively supports 12 languages and 20+ fonts.

Qwen Image 3 EditAlibaba Cloud

Qwen Image 3 Edit is Alibaba's image editing model built on the Qwen Image 3 foundation. Accepts up to 3 reference images and modifies them based on a text prompt. Supports precise edits with complex layout understanding, multilingual text rendering, and fine-grained visual detail control.

Qwen Image EditAlibaba Cloud

Qwen Image Edit is an advanced open-source image editing model by Alibaba, designed for precise and natural language-driven modifications of existing photos and graphics.

Qwen MaxAlibaba Cloud

Qwen Max: powerful AI, excels in stability, competes globally.

Qwen Max 2025-09-23Alibaba Cloud

Multilingual MoE model for diverse tasks

Qwen PlusAlibaba Cloud

Qwen-Plus: Alibaba's advanced multilingual model with enhanced reasoning and instruction-following capabilities.

Qwen Text Embedding v3Alibaba Cloud

High-quality text embedding model with support for custom dimensions (512-1024) and over 50 languages. Suitable for semantic search, recommendation systems, and text analysis.

Qwen Text Embedding v3Alibaba Cloud

High-quality text embedding model with support for custom dimensions (512-1024) and over 50 languages. Suitable for semantic search, recommendation systems, and text analysis.

Qwen Text Embedding v4Alibaba Cloud

Advanced text embedding model from the Qwen3-Embedding series with support for custom dimensions (64-2048) and over 100 languages. Ideal for semantic search, clustering, and classification tasks.

Qwen Text Embedding v4Alibaba Cloud

Qwen Text Embedding v4 is the latest iteration in Alibaba Cloud’s high-performance embedding model series, engineered to transform textual inputs into dense, semantically rich vector representations.

Qwen TurboAlibaba Cloud

Qwen Turbo: optimizes AI agent speed, integrates with RAG, large context window.

Qwen3 32BAlibaba Cloud

Open-source AI powerhouse excelling in multilingual reasoning, coding (73.9% HumanEval), and math. Features 131K context & Apache 2.0 license.

Qwen3 CoderAlibaba Cloud

Alibaba's 480B-parameter coding model — 262K-token context (extendable to 1M), built for agentic coding.

Qwen3 CoderAlibaba Cloud

Strong coding accuracy focused on software engineering tasks, but with a smaller context window and less emphasis on integration.

Qwen3 Next 80B A3B ThinkingAlibaba Cloud

Alibaba's Qwen 3 Next 80B A3B Thinking is a multilingual reasoning model with advanced chain-of-thought capabilities.

Qwen3 Thinking 2507Alibaba Cloud

Alibaba's Qwen 3 thinking 2507 is a multilingual reasoning model with knowledge augmentation, and creative capabilities.

Qwen3 TTS FlashAlibaba Cloud

Qwen3-TTS offers 17 voices and supports multiple languages and dialects including Chinese. The system can adapt its tone and smoothly process complex text.

Qwen3 VL 32B InstructAlibaba Cloud

Qwen3 VL 32B Instruct can be seamlessly integrated into multimodal applications requiring precise image-text interaction.

Qwen3 VL 32B ThinkingAlibaba Cloud

Qwen3 VL 32B Thinking is revolutionizing multimodal AI by enabling machines to process complex visual data alongside extended textual reasoning.

Qwen3 VL FlashAlibaba Cloud

Alibaba's Qwen 3 VL Flash is a fast and cost-effective vision-language model with hybrid thinking capabilities. Supports image understanding with up to 16,384 tokens per image and high-resolution mode for detailed analysis.

Qwen3 VL PlusAlibaba Cloud

Alibaba's Qwen 3 VL Plus is a powerful vision-language model with hybrid thinking capabilities. Supports image understanding with up to 16,384 tokens per image and high-resolution mode for detailed analysis.

Qwen3-Max InstructAlibaba Cloud

Qwen 3 Max Instruct sets a new benchmark for trillion-parameter language models, with massive context lengths, diverse language support, and cutting-edge performance in code and math tasks.

Qwen3-Max PreviewAlibaba Cloud

Qwen3-Max is Alibaba's trillion-parameter flagship language model designed for high-speed, large-scale tasks with an ultra-long context window.

Qwen3-Next-80B-A3B InstructAlibaba Cloud

Qwen3-Next-80B-A3B Instruct is a next-generation large language model that balances enormous parameter scale with sparse activation to deliver fast, cost-efficient, and scalable instruction-following capabilities.

Qwen3-Next-80B-A3B ThinkingAlibaba Cloud

Qwen3-Next-80B-A3B Thinking integrates seamlessly into modern AI workflows with flexible deployment options including serverless, on-demand dedicated, and reserved monthly instances.

Qwen3-Omni CaptionerAlibaba Cloud

Qwen3-Omni-Captioner automatically generates accurate and comprehensive descriptions for complex audio, including speech, ambient sounds, music, and sound effects. The model can identify speaker emotions, musical elements, and is suitable for audio content analysis, security audits, and audio editing. Supports audio input up to 40 minutes (1 second = 25 tokens).

Qwen3-TTS-Flash RealtimeAlibaba Cloud

Qwen3-TTS-Flash offers a highly responsive, robust solution for real-time and batch multilingual text-to-speech synthesis.

Qwen3.5 Omni FlashAlibaba Cloud

Qwen3.5 Omni Flash scores well above the median for non-reasoning models in its price tier on the Artificial Analysis Intelligence Index (26 vs. a median of 15), generates output at 164 tokens per second, and supports every major input modality out of the box. It's competitively priced but runs verbose, producing roughly 2.5x more output tokens than average, which affects billing on output-heavy tasks.

Qwen3.5 Omni PlusAlibaba Cloud

The Plus variant is the flagship of the Qwen3.5 Omni family, sitting above Flash (optimized for speed) and Light (optimized for edge deployment). It's an instruct model, meaning it's fine-tuned for direct use rather than raw pretraining, and it supports a 256,000-token context window, enough to hold an entire book, a feature-length film's worth of captions, or over ten hours of continuous audio.

Qwen3.5 PlusAlibaba Cloud

Qwen3.5-Plus is a commercial large language model by Alibaba Cloud designed for long-context text generation and enterprise-grade conversational AI. Supports up to 1M tokens per request with production-ready API stability.

Qwen3.5-FlashAlibaba Cloud

Built for real-time use cases, it delivers fast responses, strong reasoning for lightweight tasks, and excellent scalability, making it ideal for high-volume applications without sacrificing quality.

Qwen3.6 Max PreviewAlibaba Cloud

Qwen3.6-Max-Preview — a high-capability reasoning model by Alibaba Cloud designed for complex tasks and long-form problem solving. It provides strong reasoning performance and high accuracy for advanced use cases.

Qwen3.6-35B-A3BAlibaba Cloud

Alibaba's newest open-source mixture-of-experts model activates just 3 billion parameters at inference time, yet it goes toe-to-toe with dense models four to nine times its active size on agentic coding, reasoning, and multimodal tasks.

Qwen3.6-FlashAlibaba Cloud

It is engineered to deliver near-instant responses while maintaining solid language understanding, making it a practical choice for high-traffic applications and interactive systems.

Qwen3.6-PlusAlibaba Cloud

Qwen3.6-Plus is a general-purpose large language model developed by Alibaba Cloud, designed to strike a practical balance between performance, cost efficiency, and speed.

Qwen3.7 PlusAlibaba Cloud

Qwen3.7-Plus is a multimodal agent model that unifies vision and language into a single agent foundation. It delivers strong vision-language capabilities while retaining agentic strength in coding, tool use, and productivity workflows — including GUI understanding, visual Q&A, and full-modality input for complex software engineering tasks.

Qwen3.8 2.4T A95BAlibaba Cloud

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. Reasoning is always on.

Qwen3.8 MaxAlibaba Cloud

Qwen3.8-Max is Alibaba's flagship reasoning and agentic LLM, positioned as its highest-capability tier for complex multi-step reasoning, coding and autonomous workflows. Natively multimodal with a 1M-token context window and hybrid thinking mode.

Qwen3.8-FlashAlibaba Cloud

Alibaba's latest multimodal model, tuned for stronger coding and office-task performance while cutting training cost versus its predecessor. 262K-token context by default, expandable to 1M.

Wan 2.1Alibaba Cloud

Video model for cinematic outputs

Wan 2.1 PlusAlibaba Cloud

Wan2.1 Plus by Alibaba is an advanced text-to-video AI model optimized for generating high-quality 720P videos from text prompts.

Wan 2.1 TurboAlibaba Cloud

Wan2.1-T2V-Turbo is efficient text-to-video AI model designed for fast, high-quality video generation from textual input.

Wan 2.2 14B Animate MoveAlibaba Cloud

Bring any still photo to life. Wan 2.2 Animate Move transfers full-body movement and facial expressions from a reference video onto a static character image, producing fluid, identity-preserving HD animation at 24 fps without a single keyframe.

Wan 2.2 14B Animate ReplaceAlibaba Cloud

Wan 2.2 14B Animate Replace delivers state-of-the-art character replacement in videos.

Wan 2.2 FlashAlibaba Cloud

Wan 2.2 Flash excels in fast, high-quality image generation from text prompts.

Wan 2.2 PlusAlibaba Cloud

Wan 2.2 Plus is a cutting-edge text-to-image generation model focused on delivering enhanced stability, superior detail, and elevated artistic quality.

Wan 2.2 Plus Image to VideoAlibaba Cloud

Wan2.2 I2V by Alibaba-Cloud is a highly capable AI model designed for vision-language understanding, multi-modal reasoning, and intelligent content generation. It supports large-context multi-turn interactions with enhanced vision-to-text comprehension and generation precision.

Wan 2.2 Plus Text to VideoAlibaba Cloud

Wan2.2 T2V balances powerful multi-modal AI performance with real-world limitations in image-text understanding and processing.

Wan 2.2 Vace DepthAlibaba Cloud

Wan 2.2 Vace Depth offers a specialized solution for video creators demanding high control over spatial depth information in video generation.

Wan 2.2 Vace InpaintingAlibaba Cloud

Wan 2.2 VACE stands out as a professional-grade, multimodal video inpainting model designed to empower content creators with powerful editing and enhancement tools.

Wan 2.2 Vace OutpaintingAlibaba Cloud

Wan 2.2 VACE offers professional-grade video outpainting capabilities with versatile control modes and high-resolution output options.

Wan 2.2 Vace PoseAlibaba Cloud

Wan 2.2 Vace Pose is a cutting-edge model for controlled video generation using state-of-the-art VACE technology.

Wan 2.2 Vace ReframeAlibaba Cloud

Wan 2.2 VACE Reframe is a cutting-edge video generation and editing model focused on delivering high-quality, temporally coherent video outputs with detailed user control over multiple aspects of video content.

Wan 2.5 Image-to-Video PreviewAlibaba Cloud

Wan 2.5 marks a significant leap in AI video generation technology, blending high-quality photorealistic visuals, synchronized audio, and advanced cinematic controls.

Wan 2.5 PreviewAlibaba Cloud

Wan 2.5 Preview excels at interpreting complex prompts with precise semantic compliance, ensuring outputs closely match creative intent.

Wan 2.5 Text-to-Video PreviewAlibaba Cloud

Generate high-quality videos from text prompts using Alibaba Cloud's WAN 2.5 text-to-video model with support for 480p, 720p, and 1080p resolutions.

Wan 2.6Alibaba Cloud

Wan 2.6 delivers superior coherence, semantic alignment, and visual fidelity, ideal for commercial, creative, and technical applications.

Wan 2.6 ImageAlibaba Cloud

Wan 2.6 is a next-generation text-to-image model from Alibaba Cloud with enhanced visual quality and detail.

Wan 2.6 Image to Video FlashAlibaba Cloud

Wan 2.6 Image-to-Video Flash is a faster variant of Wan 2.6 I2V for rapid video generation.

Wan 2.6 Image-to-VideoAlibaba Cloud

Wan 2.6 Image-to-Video animates still images into smooth, high-quality videos with advanced motion modeling.

Wan 2.6 Reference-to-VideoAlibaba Cloud

WAN 2.6 is a cutting-edge reference-guided video generation model that synthesizes high-fidelity, temporally coherent videos from a single reference image and textual prompts.

Wan 2.6 Text to VideoAlibaba Cloud

Wan 2.6 Text-to-Video generates high-quality videos from text descriptions with advanced motion modeling.

Wan 2.6 VideoAlibaba Cloud

With Wan 2.6, creators can move from idea to finished video without traditional filming, editing, or animation pipelines.

Wan 2.7Alibaba Cloud

Wan 2.7 is Alibaba Tongyi Lab's most capable video generation system to date. It collapses four distinct creation modes — text-to-video, image-to-video, reference-to-video, and natural-language video editing.

Wan 2.7Alibaba Cloud

The model enables precise control over style, layout, and branding while maintaining high-quality, coherent outputs across large image sets.

Wan 2.7 ImageAlibaba Cloud

Wan 2.7 is a next-generation text-to-image model from Alibaba Cloud with enhanced visual quality and detail.

Wan 2.7 Image to VideoAlibaba Cloud

Wan 2.7 I2V — an image-to-video model by Alibaba Cloud that generates videos from input images. It supports first-frame, first-and-last-frame, and continuation workflows.

Wan 2.7 ProAlibaba Cloud

It supports 4K output, image editing, image sets, and multi-reference generation, while the broader Wan 2.7 stack is designed to help teams ship faster with simpler integration paths and more predictable operations.

Wan 2.7 Pro ImageAlibaba Cloud

Wan 2.7 Pro is a next-generation text-to-image model from Alibaba Cloud with enhanced visual quality and detail.

Wan 2.7 Reference to VideoAlibaba Cloud

Wan 2.7 R2V — a reference-to-video model by Alibaba Cloud that generates videos using reference images or videos. It enables character, motion, and style transfer.

Wan 2.7 Text to VideoAlibaba Cloud

Wan 2.7 T2V — a text-to-video model by Alibaba Cloud that generates multi-shot videos from prompts. It supports cinematic visuals, prompt control, and audio integration.

Wan 3.0 VideoAlibaba Cloud

Wan 3.0 is Alibaba Cloud's all-in-one reference-based video generation model, supporting text-to-video, image-to-video, and reference-based generation from images, video, audio, documents, and web links, up to 30 seconds long.

Z-Image TurboAlibaba Cloud

Z-Image Turbo delivers photorealistic images from text prompts using a compact 6B parameter architectureI, optimized for speed on consumer hardware.

Z-Image Turbo LoRAAlibaba Cloud

Z-Image Turbo LoRA is a highly efficient text-to-image model that delivers photorealistic images with ultra-low latency.