1# fal.ai Models Reference
2 
3## Image-to-Video
4 
5### Kling O3 Standard (RECOMMENDED)
6- Model: `fal-ai/kling-video/o3/standard/image-to-video`
7- Docs: https://fal.ai/models/fal-ai/kling-video/o3/standard/image-to-video/llms.txt
8- Price: **$0.084/s** (no audio), **$0.112/s** (with audio)
9- Duration: `"3"` to `"15"` (string)
10- Params: `image_url`, `prompt`, `duration`, `generate_audio`, `end_image_url`, `negative_prompt`, `enable_safety_checker`
11- Best balance of quality and price
12 
13### Kling v3 Pro
14- Model: `fal-ai/kling-video/v3/pro/image-to-video`
15- Docs: https://fal.ai/models/fal-ai/kling-video/v3/pro/image-to-video/llms.txt
16- Price: **$0.112/s** (no audio), **$0.168/s** (audio), **$0.196/s** (audio+voice)
17- Duration: `"3"` to `"15"` (string)
18- Params: same as O3 plus `elements`, `negative_prompt`, `cfg_scale` (0-1)
19- Supports `elements` for character/video references (max 8s with elements)
20 
21### Kling v2.5 Turbo Pro (cheapest)
22- Model: `fal-ai/kling-video/v2.5-turbo/pro/image-to-video`
23- Docs: https://fal.ai/models/fal-ai/kling-video/v2.5-turbo/pro/image-to-video/llms.txt
24- Price: **$0.07/s**
25- Fastest generation, lowest cost
26 
27### Seedance 1.5 Pro (ByteDance)
28- Model: `fal-ai/bytedance/seedance/v1.5/pro/image-to-video`
29- Docs: https://fal.ai/models/fal-ai/bytedance/seedance/v1.5/pro/image-to-video/llms.txt
30- Price: ~**$0.26 per 5s** 720p with audio
31- Duration: STRING `"4"` to `"12"` (NOT integer!)
32- Resolution: `"480p"`, `"720p"`, `"1080p"`
33- Aspect ratio: `"21:9"`, `"16:9"`, `"4:3"`, `"1:1"`, `"3:4"`, `"9:16"`, `"auto"`
34- Generates audio by default, supports `end_image_url`
35 
36## Motion Control
37 
38### Kling v3 Pro Motion Control
39- Model: `fal-ai/kling-video/v3/pro/motion-control`
40- Docs: https://fal.ai/models/fal-ai/kling-video/v3/pro/motion-control/llms.txt
41- Price: **$0.126/s**
42- Required: `image_url`, `video_url`, `character_orientation` (`"image"` or `"video"`)
43- Optional: `prompt`, `keep_original_sound`, `elements`
44- `video` orientation: up to 30s, better for complex motion
45- `image` orientation: up to 10s, better for preserving framing
46 
47### Kling v3 Standard Motion Control
48- Model: `fal-ai/kling-video/v3/standard/motion-control`
49- Docs: https://fal.ai/models/fal-ai/kling-video/v3/standard/motion-control/llms.txt
50- Same params, lower cost
51 
52## Speaking Avatar
53 
54### HeyGen Avatar4
55- Model: `fal-ai/heygen/avatar4/image-to-video`
56- Docs: https://fal.ai/models/fal-ai/heygen/avatar4/image-to-video/llms.txt
57- Price: **$0.10/s**
58- Params: `image_url`, `prompt` (speech text), `talking_style` (`"stable"` / `"expressive"`), `resolution`, `aspect_ratio`, `voice`, `audio_url`
59- Resolution: `"360p"` / `"480p"` / `"540p"` / `"720p"` / `"1080p"`
60- Aspect ratio: `"16:9"` / `"9:16"` / `"1:1"`
61 
62## Image Edit / Upscale
63 
64### Seedream v5 Lite
65- Model: `fal-ai/bytedance/seedream/v5/lite/edit`
66- Docs: https://fal.ai/models/fal-ai/bytedance/seedream/v5/lite/edit/llms.txt
67- Price: **$0.035/image**
68- Default output: `auto_2K`
69- Params: `image_urls` (list, up to 10), `prompt`, `image_size`, `enable_safety_checker`
70 
71### Seedream v4.5
72- Model: `fal-ai/bytedance/seedream/v4.5/edit`
73- Docs: https://fal.ai/models/fal-ai/bytedance/seedream/v4.5/edit/llms.txt
74- Price: **$0.04/image**
75- Output: up to 4096x4096
76- Multi-image input: can reference multiple images (Figure 1, Figure 2, etc.)
77- Great for replacing characters while preserving pose
78 
79## xAI Grok Imagine Video
80 
81### Grok Imagine Text to Video
82- Model: `xai/grok-imagine-video/text-to-video`
83- Docs: https://fal.ai/models/xai/grok-imagine-video/text-to-video/api
84- Params: `prompt`, `duration`, `aspect_ratio`, `resolution`
85- Resolution: `480p`, `720p`
86- Aspect ratio: `16:9`, `4:3`, `3:2`, `1:1`, `2:3`, `3:4`, `9:16`
87- No documented safety-checker toggle in the fal schema
88 
89### Grok Imagine Image to Video
90- Model: `xai/grok-imagine-video/image-to-video`
91- Docs: https://fal.ai/models/xai/grok-imagine-video/image-to-video/api
92- Params: `prompt`, `image_url`, `duration`, `aspect_ratio`, `resolution`
93- Good fit when you want to pull the first frame from a reference video and animate from that still
94- No documented `generate_audio`, `negative_prompt`, `end_image_url`, or safety-checker toggle in the fal schema
95 
96### Grok Imagine Edit Video
97- Model: `xai/grok-imagine-video/edit-video`
98- Docs: https://fal.ai/models/xai/grok-imagine-video/edit-video/api
99- Params: `prompt`, `video_url`, `resolution`
100- Input video is resized to a maximum area of `854x480` and truncated to `8` seconds according to the fal schema
101- Best fit for transforming an existing clip; not a dedicated "extend video timeline" endpoint
102 
103## Content Policy
104 
105- `enable_safety_checker: False` disables client-side filter only
106- Server-side content filter remains active on all models
107- For Grok Imagine Video on fal.ai, the public schema does not currently expose `enable_safety_checker`, so there is nothing explicit to disable from the client side
108- Blocked words in suggestive context: explicit anatomy, "squeezes", "revealing", "sensual"
109- Safe alternatives: "decollete", "fashion editorial", "adjusts", "touches", "graceful movements"
110 
111## Orientation Guide (Motion Control)
112 
113- `"video"`: dancing, full-body motion, complex movement, up to 30s
114- `"image"`: preserve still image framing/pose, up to 10s
115- Match the source image pose to the first frame of the video for best results
116