Wan 2.7 by Alibaba. Text-to-video at up to 1080p, 2-15 seconds, five aspect ratios, optional synced audio, and built-in prompt expansion.
Models
All Models
Wan 2.7 Image Pro by Alibaba generates up to 4K images from text and edits or fuses up to 9 reference images, with optional thinking mode for deeper prompt reasoning.
Wan 2.7 Image by Alibaba: text-to-image and multi-reference editing at 1K or 2K, with image set mode and thinking mode for up to four outputs per run.
Wan 2.7 I2V by Alibaba animates images into 720p or 1080p video, 2-15s. Supports first/last frame bracketing, clip continuation, and optional audio input.
Minimax Speech 2.8 Turbo by MiniMax. Fast TTS with 17 voices, 10 emotions, and 40+ language boosts, built for low-latency apps.
Minimax Speech 2.8 HD by MiniMax. Premium TTS with 17 voices, 10 emotions, 40+ languages, natural interjections, and precise control over speed, pitch, and volume.
Hy Wu Edit by Tencent. Transfer outfits, swap faces, and blend textures using up to 3 reference photos, with no fine-tuning required.
Google Lyria 3 Clip by Google generates 30-second music clips from text prompts or reference images, with control over genre, tempo, mood, and song structure.
Grok Edit Video by xAI: edit any clip with a text prompt. Restyle scenes, swap objects, change environments, all while preserving what you don't touch.
Grok Extend Video by xAI continues your clip from its last frame, generating 2 to 10 seconds of new AI footage guided by a text prompt.
Flux.2 LoRA for stylized 3D costume sets. Renders outfit collections and equipment on invisible figures with detailed textures and studio lighting.
Tada 3B by Hume AI clones any voice from a short audio reference and synthesizes multilingual speech across 10 languages with no transcript hallucinations.
Tada 1B by Academia / Open Source clones any voice from a short audio sample and synthesizes multilingual speech across 10 languages with speed control.
HeyGen Avatar 4 by HeyGen: turn a face photo into a talking video with 80+ preset voices, resolutions up to 1080p, and stable or expressive motion.
HeyGen Video Translate Speed dubs your video into 170+ languages with AI lip sync, optimized for fast, high-volume translation at scale.
HeyGen Video Translate Precision by HeyGen. Translate video speech into 170+ languages with high-fidelity lip sync, voice cloning, and multi-speaker support.
HeyGen Video Agent by HeyGen turns a text prompt into a polished presenter video, handling scripting, avatar selection, scenes, and narration pacing automatically.
Pixelcut Background Removal by Pixelcut: isolate subjects from any image with clean edge detection. Export as full RGBA composite or alpha-only mask.
Tripo P1 Multi View by Tripo AI. Generate high-fidelity 3D meshes with PBR maps from up to four reference angles using native 3D diffusion.
Physic Edit applies physics-aware transformations to images: flood cities, melt armor, shatter glass, or freeze scenes with accurate refraction and deformation.
A Flux 2 LoRA for 3D low-poly environments suited to platformers and adventure games, with geometric shapes, vibrant colors, and soft lighting.
Flux.2 LoRA for stylized game icons: gemstones, crystals, and enchanted loot rendered with vibrant colors and polished digital painting.
Generate editable SVG glyphs that match your brand's typography. Upload up to 8 reference characters, type your target, get a clean vector output up to 4K.
VecGlypher generates editable SVG glyphs from a text style description. Control fill, stroke color and width, and output size up to 4096 px.
LTX 2.3 Pro Retake by Lightricks. Regenerate any section of a video: replace audio, video content, or both. Define the start time and duration, prompt the change. 1080p.
LTX 2.3 Pro Extend Video by Lightricks adds new footage to the start or end of any clip. Guide the continuation with a prompt and choose 6, 8, or 10 seconds.
LTX 2.3 Fast by Lightricks: speed-optimized video generation with synchronized audio, up to 4K, 6-20s clips, portrait or landscape support, and camera motion controls.
Trellis 2 by Microsoft retextures any 3D mesh using a reference image, producing PBR-quality textures at up to 4K with adjustable guidance and resolution.
Kling V3 Pro by Kuaishou. Transfer motion from any reference video onto a character image, with facial consistency and optional audio preservation.
Kling V3 Standard Motion Control by Kuaishou. Animate any character image using a reference video to drive their movements, with optional audio transfer.
Tripo 3.1 Multi View by Tripo AI generates accurate, PBR-ready 3D meshes from up to four reference angles: front, left, back, and right.
Tripo Rigging 2.5 by Tripo AI. Auto-rig any character or creature mesh, from bipeds to serpentines, and optionally retarget with preset animations.
Vidu Q2 Pro by Shengshu Technology generates videos from up to 7 reference images or 2 reference videos, preserving subject identity at up to 1080p.
Tencent UV Unwrapping by Tencent automatically generates clean UV maps for FBX, OBJ, and GLB models up to 30,000 faces, ready for texturing.
Tencent Texture Edit by Tencent. Retexture FBX models using a text prompt or reference image. Prompt mode outputs full PBR maps for game-ready assets.
Qwen Edit Plus by Alibaba. Edit images from plain-language instructions: change objects, styles, or in-image text. Supports LoRA fine-tuning and up to 2048x2048 output.
Qwen Edit 2511 by Alibaba edits images from plain text instructions. Add, remove, or restyle elements, with up to 6 LoRA styles and multi-reference image support.
Qwen Edit 2509 by Alibaba edits images from text instructions, with multi-image input, up to 6 LoRA styles, and precise portrait and text editing.
FLUX.2 Klein 9b by Black Forest Labs. Fast text-to-image and image editing with strong style coherence, multi-reference support, and LoRA compatibility.
Vidu Q3 Turbo by Shengshu Technology animates images into 1080p video up to 16 seconds, with start/end frame control and optional background audio.
Vidu Q3 Pro by Shengshu Technology animates a single image or interpolates between a start and end frame, up to 16 seconds at 1080p.
Vidu Q3 Turbo T2V by Shengshu Technology. Fast text-to-video generation up to 16 seconds and 1080p, with optional background music and flexible aspect ratios.
Vidu Q3 Pro by Shengshu Technology. Premium text-to-video up to 16 seconds in up to 1080p, with cinematic camera control and optional audio.
Vidu Q2 by Shengshu Technology generates text-to-video clips up to 10 seconds in up to 1080p, with cinematic camera control and optional background music.
Vidu Q2 Reference2V by Shengshu Technology generates videos with up to 7 reference images to maintain consistent characters, props, and costumes across every frame.
Vidu Q2 Turbo I2V by Shengshu Technology. Fast image-to-video with smooth motion, first/last-frame control, up to 1080p and 10 seconds.
Vidu Q2 Pro Fast by Shengshu Technology animates still images into video at speed, with 720p or 1080p output, up to 10 seconds, and optional background music.
Vidu Q2 Pro I2V by Shengshu Technology animates a single image or bridges two frames into cinematic video up to 1080p, with strong subject fidelity.
Generate 5-second 1080p videos from up to 7 reference images. Vidu Q1 preserves subject appearance, style, and texture through motion. 16:9, 1:1, or 9:16.