Alibaba models on AIgateway — pricing, context, capabilities

Create custom voices using Qwen3-TTS Voice Design model and later use Clone Voice model to create your own voices!

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks.

Alibaba's Wan 2.6 text-to-image model generating images from text prompts with optional negative prompts and customizable dimensions.

Qwen-Image is an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing.

Qwen-Image-2.0 is a next-generation foundational unified generation-and-editing model

Qwen Image 2512 is an improved version of Qwen Image with better text rendering, finer natural textures, and more realistic human generation.

LoRA inference endpoint for Qwen Image 2512, an improved version of Qwen Image with better text rendering, finer natural textures, and more realistic human generation.

Endpoint for Qwen's Image Editing Plus model also known as Qwen-Image-Edit-2509. Has superior text editing capabilities and multi-image support.

Generate full portrait from a cropped face photo

Removes harsh shadows and light spots from images, replacing them with soft, even, natural-looking illumination.

Precise camera position and angle control (rotation, zoom, vertical movement)

Remove unwanted elements (objects, people, text) while maintaining image consistency

Endpoint for Qwen's Image Editing 2511 model.

Generates same scene from different angles (azimuth/elevation) with Qwen image Edit 2511 and the Lora Multiple Angles

LoRA inference endpoint for the Qwen Image Editing model.

Add a realistic scene behind the object with white background

Create group photos

Create cinematic transitions and scene progressions (camera movements, framing changes)

Remove unwanted elements (objects, people, text) while maintaining image consistency

Remove existing lighting and apply soft, even illumination

Image to Image Endpoint for Qwen's Image Editing model. Has superior text editing capabilities.

Inpainting Endpoint for the Qwen Edit Image editing model.

Qwen-Image-Layered is a model capable of decomposing an image into multiple RGBA layers.

Qwen-Image-Layered is a model capable of decomposing an image into multiple RGBA layers. Use loras to get your custom outputs.

Wan 2.2's 5B model generates high-resolution, photorealistic images with powerful prompt understanding and fine-grained visual detail

Wan 2.2's 14B model edit high-resolution, photorealistic images with powerful prompt understanding and fine-grained visual detail

Wan 2.2's 14B model generates high-resolution, photorealistic images with powerful prompt understanding and fine-grained visual detail

Transform and edit existing images with text-guided instructions using the WAN 2.7 model for creative image manipulation.

Generate high-quality images from text prompts using the WAN 2.7 model with advanced prompt understanding and detailed output.

Z-Image is the foundation model of the Z- Image family, engineered for good quality, robust generative diversity, broad stylistic coverage, and precise prompt adherence.

Generate images from text and edge, depth or pose images using Z-Image Turbo, Tongyi-MAI's super-fast 6B model.

Generate images from text and images using Z-Image Turbo, Tongyi-MAI's super-fast 6B model.

Generate images from text and images using custom LoRA and Z-Image Turbo, Tongyi-MAI's super-fast 6B model.

Generate images from text, an image, a mask and custom LoRA using Z-Image Turbo, Tongyi-MAI's super-fast 6B model.

Generate seamlessly tiling photorealistic images from text using Z-Image Turbo

Clone your voices using Qwen3-TTS Clone-Voice model with zero shot cloning capabilities and use it on text-to-speech models to create speeches of yours!

QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.

Alibaba's Qwen 3.5 is a 397B-parameter mixture-of-experts model with 17B active parameters, offering strong reasoning capabilities with efficient inference.

Alibaba's Qwen 3 Max is a large language model with strong coding, reasoning, and multilingual capabilities, served via DashScope's OpenAI-compatible endpoint.

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support.

Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). As of now, Qwen2.5-Coder has covered six mainstream model sizes, 0.5, 1.5, 3, 7, 14, 32 billion parameters, to meet the needs of different developers. Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:

Alibaba's HappyHorse 1.1 text-to-video model. Generates videos from a text prompt with stronger dynamic expressiveness, better visual quality, and improved instruction following over 1.0. Configurable resolution, aspect ratio, and duration (3-15s).

Alibaba's HappyHorse 1.1 image-to-video model. Animates a reference image with an optional text prompt, with smoother motion, natural skin textures, and improved close-up quality over 1.0. Supports 720P and 1080P output with durations from 3 to 15 seconds.

Alibaba's HappyHorse 1.1 reference-to-video model. Takes 1-9 reference images (characters and scenes) and a prompt that choreographs them into a single video, keeping each subject's identity consistent. Supports 720P and 1080P output with durations from 3 to 15 seconds.

Alibaba's Wan 2.7 image-to-video model that generates videos from a reference image with optional text prompts. Supports 720P and 1080P output with durations from 2 to 15 seconds.

Alibaba's HappyHorse 1.0 image-to-video model. Animates a reference image with an optional text prompt. Supports 720P and 1080P output with durations from 3 to 15 seconds.

Alibaba's HappyHorse 1.0 text-to-video model. Generates videos from a text prompt with configurable resolution, aspect ratio, and duration (3-15s).

Generate 1080p video with synchronized native audio from a text prompt and references. Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4. Duration: 3–15s.

Generate 1080p video with synchronized native audio from a text prompt. Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4. Duration: 3–15s.

Happy Horse 1.1 is Alibaba's #1-ranked video model. This reference-to-video endpoint turns up to 9 reference images into 1080p video with synchronized native audio and multilingual lip-sync for consistent characters.

VACE Fun for Wan 2.2 A14B from Alibaba-PAI

VACE Fun for Wan 2.2 A14B from Alibaba-PAI

Wan 2.5 image-to-video model.

Wan 2.5 text-to-video model.

Wan-2.1 is a image-to-video model that generates high-quality videos with high visual quality and motion diversity from images

Add custom LoRAs to Wan-2.1 is a image-to-video model that generates high-quality videos with high visual quality and motion diversity from images

Add custom LoRAs to Wan-2.1 is a text-to-video model that generates high-quality videos with high visual quality and motion diversity from images

VACE is a video generation model that uses a source image, mask, and video to create prompted videos with controllable sources.

VACE is a video generation model that uses a source image, mask, and video to create prompted videos with controllable sources.

VACE is a video generation model that uses a source image, mask, and video to create prompted videos with controllable sources.

Reframe entire videos scene-by-scene using Wan VACE 2.1

Wan-Animate Replace is a model that can integrate animated characters into reference videos, replacing the original character while preserving the scene’s lighting and color tone for seamless environmental integration.

Wan 2.2's 5B model produces up to 5 seconds of video 720p at 24FPS with fluid motion and powerful prompt understanding

Wan 2.2's 5B model produces up to 5 seconds of video 720p at 24FPS with fluid motion and powerful prompt understanding

Wan-2.2 text-to-video is a video model that generates high-quality videos with high visual quality and motion diversity from text prompts.

Wan 2.7 is the latest generation AI video model, delivering enhanced motion smoothness, superior scene fidelity, and greater visual coherence.

Wan 2.6 image-to-video flash model.