Conceitos
Modelos
O catálogo completo. Cada entrada é um slug que você passa como model em POST /generations — sejam modelos de mídia ou tools de pós-produção.
GET /models devolve este mesmo resumo (slug, nome, tipo, categoria, descrição); GET /models/:slug devolve o schema completo de campos usado em cada página abaixo.
Imagem image 24
Flux Kontext Max
Flux Kontext Max — premium tier of Kontext family, with optional in-context editing from 1 reference.
Flux Kontext Pro
Flux Kontext Pro — text-to-image with optional in-context editing from 1 reference. Ratios up to 21:9.
FLUX.2 Flex
FLUX.2 Flex — flexible speed/quality, with optional reference editing (up to 8). 1K/2K.
FLUX.2 Pro
Black Forest Labs FLUX.2 Pro — premium typography & editorial, with optional reference editing (up to 8). 1K/2K.
GPT Image 1.5
OpenAI GPT Image 1.5 — legacy create with optional reference editing + mask.
GPT Image 2
OpenAI GPT Image 2 — text-to-image with optional references. Quality low/medium/high; 1K/2K/4K. Returns one image per request.
GPT-4o Image
OpenAI GPT-4o Image — legacy create with optional reference editing + mask. 1:1/2:3/3:2.
Grok Imagine Image
xAI Grok Imagine — text/image-to-image. Aspect 1:1 / 2:3 / 3:2 / 16:9 / 9:16.
Grok Imagine Quality
xAI Grok Imagine Quality — geração e edição em 1K/2K com até 3 imagens de referência.
Kling O1 Image Edit
Kling O1 Image — high-consistency edit. Edit/Variation only (NOT Create). 1-10 refs, n 1-9.
Kling O3 Image
Kling O3 Image — strong scene composition with optional reference editing. 1K/2K/4K. n 1-9.
Nano Banana
Gemini 2.5 Flash image — fast economy/legacy fallback with optional reference editing.
Nano Banana 2 Lite
Nano Banana 2 Lite do Google/Gemini — o modelo de imagem mais rápido e econômico para rascunhos ágeis e geração em alto volume. Texto→imagem + edição por referência.
Nano Banana Pro
Google/Gemini Nano Banana Pro — fast premium create with optional reference editing. 1K/2K/4K resolution.
Recraft Crisp Upscale
Recraft crisp upscale — nitidez em um passo + remoção de artefatos.
Recraft Remove Background
Recraft remove background — recorta o sujeito e devolve PNG com transparência.
Seedream 4
Seedream 4 — text-to-image with optional refs. 1K/2K/4K. n 1-15.
Seedream 4.5
ByteDance Seedream 4.5 — create + image-to-image. 2K/4K/custom.
Seedream 5.0 Lite
ByteDance Seedream 5.0 Lite — create + image-to-image. 2K/3K/custom.
Seedream 5.0 Pro
ByteDance Seedream 5.0 Pro — geração e edição de imagem topo de linha: prompt preciso, controle de layout e edição com múltiplas referências.
Topaz Image Upscale
Topaz Labs AI image upscale — restaura detalhes em 1×/2×/4×/8× preservando textura e nitidez.
Wan 2.7 Image
Wan 2.7 Image — unified create/edit. n 1-4.
Wan 2.7 Image Pro
Alibaba Wan 2.7 Image Pro — unified create/edit. Custom size + n 1-4.
Z-Image
Z-Image unified create/edit. Without image_urls = text-to-image. Economy choice for volume.
Vídeo video 41
Grok Imagine Video
xAI Grok Imagine — text/image-to-video. 6s/10s. Style modes (fun/normal/spicy).
Grok Imagine Video 1.5
xAI Grok Imagine 1.5 — imagem-para-vídeo. 1–15s. 480p/720p.
Hailuo 02
MiniMax Hailuo 02 Standard — text/image guided. 512p/768p.
Hailuo 02 Pro
MiniMax Hailuo 02 Pro — premium 1080p. Fixed per-video price.
Hailuo 2.3
MiniMax Hailuo 2.3 — text-to-video with optional first-frame image. 768p (6s/10s) or 1080p (6s only).
Happy Horse 1.0
Alibaba Happy Horse — text/image/reference. 720p/1080p. 3–15s.
Happy Horse 1.0 Video Edit
Source video + prompt edit. Optional reference imagery. 720p/1080p, source 3–60s, billing cap 15s.
Happy Horse 1.1
Alibaba Happy Horse 1.1 — texto→vídeo, imagem→vídeo e referência→vídeo. 720p/1080p, 3–15s.
Infinitalk
InfiniteTalk — vídeo falante a partir de imagem + áudio, até 15s. 480p / 720p.
Kling 2.1
Kling 2.1 — image-to-video. 5s/10s. Pro tier adds optional end-frame and costs ~1.85x of Standard.
Kling 2.5 Turbo Pro
Kling 2.5 Turbo Pro — text + start/end frame. 5s/10s.
Kling 2.6
Kling 2.6 — text + image-to-video, optional end frame (forces sound off). 5s/10s.
Kling 2.6 Motion Control
Character animation via motion transfer. 720p/1080p.
Kling 3.0
Kling 3.0 — Standard/Pro tiers. Text + image-to-video (start + end frame OR reference), multi-shot and @elements. 3–15s.
Kling 3.0 Motion Control
Motion transfer: 1 reference image + 1 reference video. 720p/1080p.
Kling 3.0 Turbo
Kling 3.0 Turbo — otimizado para velocidade. Tiers Standard (720p) / Pro (1080p). Texto e imagem-para-vídeo (primeiro frame), multi-shot até 6 cenas. 3–15s.
Kling Avatar 2.0
Kling Avatar 2.0 — avatar que fala a partir de imagem + áudio. Standard/Pro.
Kling O3
Kling O3 — Standard/Pro. Text + image (start+end frame) + reference + @elements + multi-shot.
MiniMax H3 / Hailuo 03
MiniMax H3 (Hailuo 03) — 2K nativo, 5-15s, com referências de imagem, vídeo e áudio. Start/end frame.
Omni Flash
Google Omni Flash — texto, imagem (1 ou 3 refs) ou vídeo-para-vídeo. 720p a 4K, 4-10s.
OmniHuman 1.5
OmniHuman 1.5 — avatar falante a partir de um retrato + áudio. Pessoas, pets, anime.
Runway Gen-4.5
Runway Gen-4.5 — 5s/10s. 720p/1080p. Text + image-to-video.
Seedance 1.0 Pro
Seedance 1.0 Pro — 5s/10s. 720p/1080p.
Seedance 1.5 Pro
Seedance 1.5 Pro — start + last frame. 4s/8s/12s. 480p/720p/1080p.
Seedance 2
ByteDance Seedance 2 — text/image/reference + start-last frame. 480p–1080p. Optional audio.
Seedance 2 Fast
Seedance 2 Fast — 480p/720p only. Per-second.
Seedance 2 Mini
Seedance 2 Mini — variante econômica do Seedance 2. 480p/720p. Por segundo. Áudio opcional.
Sora 2 Official
OpenAI Sora 2 Official — fixed durations 4/8/12/16/20s.
Sora 2 Pro Official
OpenAI Sora 2 Pro Official — 720p/1024p/1080p. Text + image-to-video.
Topaz Video Upscale
Topaz Labs video upscale — MP4/MOV/MKV até 50MB, fator 1×/2×/4×.
Veo 3.1 Fast
Google Veo 3.1 Fast — 8s fixed. 720p/1080p/4K. Text + image-to-video + frame.
Veo 3.1 Lite
Google Veo 3.1 Lite — 8s. Text-only. 720p/1080p/4K.
Veo 3.1 Quality
Google Veo 3.1 Quality — 8s. 720p/1080p/4K.
Volcengine Lip Sync
Volcengine lip sync vídeo-para-vídeo — re-sincroniza um vídeo existente com um novo áudio.
Wan 2.2 Fast
Wan 2.2 Fast — text-to-video.
Wan 2.5 Video
Wan 2.5 — text-to-video. 5s/10s.
Wan 2.6 Video
Wan 2.6 — text-to-video, multi-shot. 5/10/15s. 720p/1080p.
Wan 2.6 Video-to-Video
Wan 2.6 V2V edit/restyle. 5s/10s. 720p/1080p.
Wan 2.7 Video
Alibaba Wan 2.7 — text-to-video. 720p/1080p. Per-second pricing.
Wan Animate Move
Animate target image using motion from source video. 480p/580p/720p.
Wan Animate Replace
Replace character/object using source motion video + target image. 480p/580p/720p.
Áudio audio 21
Add Instrumental
Generate an instrumental track from a vocal or text prompt.
Add Vocals
Add lead/backing vocals to an instrumental track.
AI Music Video
Generate a simple visualizer/music-video for a track.
Boost Music Style
Take a short style description and expand it into a richer prompt.
Convert to WAV
Convert a generated track to WAV format for delivery.
ElevenLabs Music
ElevenLabs Music — trilhas completas por prompt, 3s a 10min, instrumental ou com vocais.
ElevenLabs TTS Turbo v2.5
ElevenLabs Turbo v2.5 — TTS multilíngue de baixa latência pela metade do preço do V3.
ElevenLabs V3 TTS
ElevenLabs V3 text-to-speech — narração multilíngue de alta fidelidade com 21 vozes.
Extend Music
Continue an existing Suno song from a chosen timestamp. Same model_version as the source clip.
Gemini 3.1 Flash TTS
Gemini 3.1 Flash TTS — narração com direção de estilo em linguagem natural, 30 vozes e 70+ idiomas.
Generate Lyrics
Generate song lyrics from a topic, mood and style.
Generate Music Cover
Generate a cover-art image for a generated song.
Generate Persona
Generate a persona/voice/style identity for reuse across tracks.
Get Timestamped Lyrics
Return aligned lyrics with timestamps for an existing Suno song.
MiniMax Music 2.6
MiniMax Music 2.6 — alternative music generator. High-fidelity vocals and instrumentation, multilingual support.
Replace Section
Replace a section of an existing track from t1..t2 with newly generated content.
Suno Music
Suno AI music generator — high-quality songs with vocals, chorus and structure. Lyrics auto-written or supplied.
Upload and Cover Audio
Upload a reference clip and let Suno generate a cover/remake in a chosen style.
Upload and Extend Audio
Upload audio and extend it into a longer track.
Vocal Remover
Separate vocals and instrumental from a source track.
xAI TTS 1
xAI TTS 1 — voz econômica com tags expressivas ([laugh], <whisper>) e 5 vozes.
Tools tool 6
Camera Angle
Drag a 3D camera around the subject — change rotation, tilt and zoom while keeping the subject, environment and lighting identical.
Enhancer
Enhance image quality and details with curated one-click presets.
Inpaint
Paint a mask over the area you want to change and describe the edit. Optionally attach a reference image (Image3) for the model to integrate.
Relight
Relight an image: change direction, quality, brightness and color of the light source while preserving the subject.
Remover Fundo
Remove o fundo da imagem automaticamente e devolve um PNG transparente (Recraft).
Upscale
Aumente a resolução de imagens e vídeos com vários modelos especializados (Topaz, Recraft).