Trending models
Start, make and scale your ideas with hundreds of workflows and every top model that gets it done.
Image models
- GPT Image 2.6 SunburstOpenAI's most detailed GPT Image 2.6 tier, built for intricate detail and editing precision; takes text and image input and outputs images in low, medium, high, xhigh, or max quality.
- GPT Image 2.6 FlareOpenAI's fast GPT Image 2.6 tier for everyday image generation and editing, with natural lighting, rich textures and complex layouts; takes text and image input and outputs images in low, medium, high, xhigh, or max quality.
- Nano Banana 2.1Latest high-efficiency Nano Banana. Better realism, prompt adherence, text rendering and character consistency at Flash speed. Great for campaigns, infographics, and iterative edits.Templates
- GPT Image 2.5 SunburstOpenAI's most capable GPT Image 2.5 tier, built for editing precision; takes text and image input and outputs images in low, medium, high, xhigh, or max quality.Templates
- GPT Image 2.5 FlareOpenAI's fastest GPT Image 2.5 tier for high-quality everyday image generation and editing; takes text and image input and outputs images.Templates
- FLUX 2 ProBlack Forest Labs FLUX.2 [pro] text-to-image and multi-reference editing model; takes a prompt and up to 8 input images and outputs 1K images.
- FLUX 3 ImageBlack Forest Labs FLUX 3 Image text-to-image and image-editing model; takes a prompt plus up to 10 reference images and generates up to 4K, with multilingual text rendering.
- FLUX 3 Image KleinBlack Forest Labs FLUX 3 Image Klein, the fast open-weights FLUX 3 model; text-to-image and editing with up to 10 reference images.
- GPT Image 2The most capable model for image editing and complex creative instructions. Excels at text rendering, visual consistency, design generation, and iterative refinement.Templates
- Grok Imagine Image 2.0xAI Grok Imagine Image 2.0 text-to-image and editing; takes a prompt and up to 5 images and outputs 1K or 2K in low or medium quality.
- Ideogram 4.5Ideogram's precise edit model: repeated edits keep the untouched pixels and colours intact, at the source's native resolution. Also generates from text.Templates
- Kling Image 3.0Kling Image 3.0 text-to-image and single-image restyling; takes a prompt and an optional reference image and outputs 1K or 2K.
- Kling Image O3Kling Image O3 text-to-image and multi-reference image editing; takes a prompt and up to 10 reference images and outputs 1K, 2K or 4K.
Krea 3Krea 3 photoreal image generation and multi-image editing; takes a prompt and up to 8 reference images and outputs up to 4 megapixels.- Meta MuseMeta's image model with built-in world knowledge and strong prompt adherence. Great for concept-rich scenes, realistic compositions, and reference-guided edits.Templates
- MiniMax H3 ImageMiniMax H3 Image text-to-image and multi-image editing with precise instruction following; takes a prompt and up to 16 reference images and outputs 1K, 2K or 4K.
- MiniMax H3 Image HighspeedMiniMax H3 Image Highspeed, the faster tier of H3 Image for text-to-image and multi-image editing; takes a prompt and up to 16 reference images and outputs 1K, 2K or 4K.
- Nano Banana 2Generate high-quality images at speed. Perfect for brainstorming, concept testing, creative exploration, and large batch workflows.Templates
- Nano Banana LiteThe fastest and most affordable image model. Ideal for rapid drafts, bulk exploration, and iterative concept testing at 1K resolution.
- Nano Banana ProCreate stunning, production-quality visuals with exceptional realism, detail, and typography. Best for campaigns, e-commerce, product photography, and hero creatives.Templates
- Qwen Image 3Alibaba Qwen Image 3 text-to-image and multi-image editing with strong text rendering; takes a prompt and up to 3 reference images and outputs 1K or 2K.
Recraft V4.1 FlashRecraft V4.1 Flash fast design-first text-to-image for brand, editorial and layout work; takes a prompt and outputs a raster image.- Seedream 5 ProByteDance Seedream 5.0 Pro text-to-image and multi-reference image-to-image (up to 10 inputs) with region editing; outputs 1K or 2K.Templates
- Seedream 5.5ByteDance Seedream 5.5 text-to-image and multi-reference editing (up to 10 inputs): edits that keep what you did not ask to change, transparent PNG output and sharper text; outputs 1K or 2K.
SenseNova U1 InfographicSenseNova U1 Infographic turns a detailed prompt into a laid-out infographic with charts, sections and legible text.- Uni-1Luma Uni-1 fast image generation and editing; takes a prompt and up to 9 reference images and outputs 2K.
- Uni-1 MaxLuma Uni-1 Max higher-quality image generation and editing; takes a prompt and up to 9 reference images and outputs 2K.
Video models
Pruna P-Video 2 ProPruna P-Video 2 Pro: text or image to video with first and last frame control, 5 to 15 seconds at 480p or 768p.- FLUX 3Black Forest Labs FLUX 3 video model with synchronized audio; text-to-video, a start frame, start and end frames, or a storyboard of up to 8 images; 5 to 20 seconds at 720p or 1080p.
- Gemini Omni 1.1 FlashGoogle Gemini Omni Flash 1.1 conversational video generation and editing; text, image, and up to 10-second video input, 3-10 second output at up to 4K and 24 FPS.Templates
- Gemini Omni 1.2 FlashGoogle Gemini Omni Flash 1.2 conversational video generation, editing and extension; text, image, and up to 15-second video input, 3-15 second output at up to 4K and 24 FPS (start and end frames up to 10 seconds), extendable to 30 seconds.
- Grok ImaginexAI's reference-to-video model. Generates short clips conditioned on up to 7 reference images with flexible aspect ratios.
- Grok Imagine 1.5xAI Grok Imagine 1.5: text- and image-to-video at native 1080p with lip-synced audio, or up to 14 reference images and pinned first and last frames at 720p, up to 15 seconds.
- Happy Horse 1.1Alibaba Happy Horse 1.1: text-to-video, image-to-video and reference-to-video from up to 9 images, with native audio and lip-sync, 3-15 seconds at up to 1080p.
- Kling 2.5Kuaishou Kling 2.5 Turbo Pro image-to-video; animates a single input image from a text prompt, 5- or 10-second clips.
- Kling 3.0Kuaishou Kling 3.0 Pro image-to-video; a required start image plus a text prompt (or multi-shot prompts) with optional end image and native audio, 3-15 second clips.
- Kling 4.0Kuaishou Kling 4.0: text-to-video, start/end frames, or reference-to-video from up to 10 images and 5 videos (30 seconds combined), up to 30 seconds at 720p or 1080p. Prompts name references @Image1, @Video1.
- Kling 4.0 FlashKuaishou Kling 4.0 Flash: fast text-to-video, a start frame, or reference-to-video from up to 10 images and 5 videos (20 seconds combined), 3 to 20 seconds at 720p. Prompts name references @Image1, @Video1.
- Kling O3Kuaishou Kling O3 Pro: text, start/end frames and up to 7 reference images in one 3-15 second clip with consistent characters and optional native audio.
- Kling V3 OmniKuaishou Kling Video 3.0 Omni generation from text, start/end images and up to 7 reference images, with optional native audio; up to 15 seconds at up to 1080p.
- MiniMax H3MiniMax H3 multimodal video model for text-to-video, image-to-video (first/last frame), and reference-to-video with native audio, up to 4K.
- MiniMax H3 MaxHigh-speed MiniMax H3 Max variant for text-to-video and image-to-video (first/last frame) only, up to 1080p.
- MiniMax H3 Max ReferenceMiniMax H3 Max reference-to-video; conditions on reference images, videos, and audio.
- MiniMax H3 Max TurboThe fast tier of MiniMax H3 Max: text-to-video and start/end-frame image-to-video with native audio and an optional soundtrack clip, up to 15 seconds at 1080p.
- MiniMax H3 ReferenceMiniMax H3 reference-to-video mode guided by up to 9 reference images, 3 reference videos, and 3 audio clips; reference and image-to-video inputs cannot be mixed.Templates
- Ray 3.2Luma Ray 3.2 text- and image-to-video with optional first and last frames, 5- or 10-second clips at up to 1080p.
- Seedance 2.0High-quality ByteDance video generation guided by reference images, for clips up to 15 seconds.Templates
- Seedance 2.0 FastSpeed- and cost-optimized ByteDance Seedance 2.0 Fast; text plus image, video, and audio references to generate or edit 4-15 second clips.Templates
- Seedance 2.5ByteDance Seedance 2.5 for long-form storytelling and omni-reference generation/editing; up to 50 reference assets and 4-30 second clips at up to 1080p.Templates
- Wan 3Alibaba Wan 3 video model for text-to-video, image-to-video (start/end frame), and reference-to-video from up to 10 images, 5 videos and 5 audio clips, with native audio; up to 30 seconds at up to 1080p.Templates
- Wan 3 PrimeAlibaba Wan 3 Prime, the faster-generating tier of Wan 3, for text-to-video, image-to-video (start/end frame), and reference-to-video from up to 10 images, 5 videos and 5 audio clips, with native audio; up to 30 seconds at up to 1080p.
Audio models
Music
ElevenLabs Music v2.5ElevenLabs Music v2.5 text-to-music; full songs or instrumentals from 3 seconds to 1 minute.- Lyria 3 ClipGoogle Lyria 3 Clip short music clips, loops, and previews; takes text and up to 10 image prompts and returns a fixed 30-second MP3 plus lyrics.
- Lyria 3 ProGoogle Lyria 3 Pro flagship music model for full-length songs with verses, choruses, and bridges; takes text and up to 10 image prompts and returns MP3 plus lyrics.
- Lyria 3.5Google Lyria 3.5 full-length songs with verses, choruses and bridges; takes text and up to 10 image prompts and returns 44.1 kHz stereo MP3 plus lyrics.
- MiniMax Music 3MiniMax Music 3 songs from a style prompt and your lyrics (or instrumental), up to 5 minutes.
Sound effects
ElevenLabs Sound Effects v2ElevenLabs Sound Effects v2; a sound effect from a text description, up to 22 seconds.
Mirelo SFX 2.0Mirelo SFX 2.0 sound effects from a text description.
Mirelo SFX 2.0 (video)Mirelo SFX 2.0 scores a video: sound effects synced to the picture, keeping its speech.
Speech
ElevenLabs Multilingual v2ElevenLabs Multilingual v2 text to speech in 29 languages.
ElevenLabs v4ElevenLabs v4 text to speech; expressive narration with audio tags, up to 5,000 characters.
ElevenLabs v4 (direct)ElevenLabs v4 text to speech with audio tags, in a premade voice or your own cloned voice; up to 10,000 characters.
ElevenLabs v4 TurboElevenLabs v4 Turbo text to speech; faster and cheaper v4 voices, up to 5,000 characters.
Fish Audio Drama 3Fish Audio Drama 3 text to speech you direct like an actor: [directions], sounds, pauses and per-word tags.
Fish Audio S2.1 ProFish Audio S2.1 Pro text to speech in 83 languages, with [bracket] directions and your own cloned voices.
Fish Audio S2.1 Pro FreeFish Audio S2.1 Pro text to speech, with the same voices as S2.1 Pro.- Gemini 3.8 Flash Lite TTSGoogle Gemini 3.8 Flash Lite text to speech; the low-cost tier with the same 30 voices.
- Gemini 3.8 Flash TTSGoogle Gemini 3.8 Flash text to speech with 30 prebuilt voices and separate style instructions.
- MiniMax Speech 2.8 HDMiniMax Speech 2.8 HD text to speech; studio-quality narration in over 300 voices, with pause and interjection tags.
Lipsync
- MiniMax H3 Max Lip SyncMiniMax H3 Max lip sync; animates a portrait image into a talking video as long as the speech, up to 2K.
Sync Lipsync 2Sync Lipsync 2 re-times the mouth in a video to new speech.
Sync Lipsync 2 ProSync Lipsync 2 Pro; higher-fidelity mouth detail and teeth than Lipsync 2.
Sync Lipsync 3Sync Lipsync 3 (sync-3) re-times a video's lips, face and head to new speech.
VEED Lipsync v2VEED Lipsync v2 re-times the mouth in a video to a new audio track.
Transcription
ElevenLabs Scribe v2ElevenLabs Scribe v2 speech to text with word timings, speaker labels and audio-event tags.

