Image-to-Video Endpoints

Curated picks across 6 use cases. Seedance 2.0 dominates for general I2V; Kling O3 / V3 specializes in reference-to-video and 4K; Sora 2 is included for I2V. Avatar/lipsync has its own large bucket. Verify with genmedia models --endpoint_id <id> --json before running.

Premium realism

Final-quality image-to-video.

• bytedance/seedance-2.0/image-to-video: ByteDance · Seedance 2.0
• bytedance/seedance-2.0/reference-to-video: ByteDance · Seedance 2.0 Reference
• fal-ai/bytedance/seedance/v1.5/pro/image-to-video: ByteDance · Seedance 1.5 Pro
• fal-ai/veo3.1/image-to-video: Google · Veo 3.1
• fal-ai/kling-video/v3/pro/image-to-video: Kling · V3 Pro
• fal-ai/kling-video/o3/pro/image-to-video: Kling · O3 Pro
• fal-ai/kling-video/o3/pro/reference-to-video: Kling · O3 Pro Reference
• fal-ai/sora-2/image-to-video: OpenAI · Sora 2
• fal-ai/minimax/hailuo-2.3/pro/image-to-video: Minimax · Hailuo 2.3 Pro

Fast / cheap

Economical / fast I2V.

• bytedance/seedance-2.0/fast/image-to-video: ByteDance
• bytedance/seedance-2.0/fast/reference-to-video: ByteDance
• fal-ai/bytedance/seedance/v1/pro/fast/image-to-video: ByteDance
• xai/grok-imagine-video/image-to-video: xAI · Grok Imagine
• xai/grok-imagine-video/reference-to-video: xAI
• fal-ai/veo3.1/lite/image-to-video: Google · Veo 3.1 Lite
• fal-ai/kling-video/v3/standard/image-to-video: Kling · V3 Standard
• fal-ai/minimax/hailuo-2.3/standard/image-to-video: Minimax

First/last frame interpolation

Controlled transition between start and end frames.

• fal-ai/kling-video/o1/image-to-video: Kling · O1 First-Last (Pro)
• fal-ai/kling-video/o1/standard/image-to-video: Kling · O1 First-Last (Standard)
• fal-ai/veo3.1/first-last-frame-to-video: Google · Veo 3.1
• fal-ai/veo3.1/fast/first-last-frame-to-video: Google · Veo 3.1 Fast
• fal-ai/veo3.1/lite/first-last-frame-to-video: Google · Veo 3.1 Lite
• fal-ai/wan-flf2v: Alibaba · Wan 2.1 First-Last
• fal-ai/vidu/q1/start-end-to-video: Vidu
• fal-ai/vidu/start-end-to-video: Vidu
• fal-ai/pixverse/c1/image-to-video: PixVerse C1

Reference-to-video

Multiple reference images (person / style / element) → video.

• bytedance/seedance-2.0/reference-to-video: ByteDance · Seedance 2.0
• bytedance/seedance-2.0/fast/reference-to-video: ByteDance · Fast
• fal-ai/bytedance/seedance/v1/lite/reference-to-video: ByteDance · Lite
• fal-ai/kling-video/o3/pro/reference-to-video: Kling · O3 Pro
• fal-ai/kling-video/o3/standard/reference-to-video: Kling · O3 Standard
• fal-ai/kling-video/o3/4k/reference-to-video: Kling · O3 4K
• alibaba/happy-horse/reference-to-video: Alibaba · Happy Horse
• xai/grok-imagine-video/reference-to-video: xAI · Grok Imagine
• fal-ai/veo3.1/reference-to-video: Google · Veo 3.1
• fal-ai/wan/v2.7/reference-to-video: Alibaba · Wan 2.7
• fal-ai/pixverse/c1/reference-to-video: PixVerse C1

4K capable

Endpoints with native 4K output.

• fal-ai/kling-video/v3/4k/image-to-video: Kling · V3 4K
• fal-ai/kling-video/o3/4k/image-to-video: Kling · O3 4K
• fal-ai/kling-video/o3/4k/reference-to-video: Kling · O3 4K Reference

Avatar / talking head / lipsync

Talking head, avatar, lip-sync video. The widest bucket in this modality, models differ meaningfully in input requirements, motion style, and language support.

• veed/fabric-1.0: Veed · Fabric 1.0 (image + audio)
• fal-ai/creatify/aurora: Creatify · Aurora
• fal-ai/bytedance/omnihuman: ByteDance · OmniHuman
• fal-ai/bytedance/omnihuman/v1.5: ByteDance · OmniHuman v1.5
• bytedance/lynx: ByteDance · Lynx
• fal-ai/kling-video/v1/standard/ai-avatar: Kling · AI Avatar (Standard)
• fal-ai/kling-video/v1/pro/ai-avatar: Kling · AI Avatar (Pro)
• fal-ai/kling-video/ai-avatar/v2/pro: Kling · AI Avatar v2 Pro
• fal-ai/kling-video/ai-avatar/v2/standard: Kling · AI Avatar v2 Standard
• fal-ai/hunyuan-avatar: Tencent · Hunyuan Avatar
• fal-ai/hunyuan-portrait: Tencent · Hunyuan Portrait
• fal-ai/hunyuan-custom: Tencent · Hunyuan Custom
• fal-ai/kling-video/lipsync/audio-to-video: Kling · Lipsync Audio-to-Video
• fal-ai/kling-video/lipsync/text-to-video: Kling · Lipsync Text-to-Video
• fal-ai/flashtalk: Flashtalk

For the multi-step TTS → lipsync recipe, see fal-recipes/references/character-lipsync.md.

Family-specific prompting

• Kling → fal-prompting/references/kling.md
• Happy Horse → fal-prompting/references/happy-horse.md

Discovery

genmedia models --category image-to-video --limit 10 --json
genmedia docs "image to video" --json