img_uuid, then poll or receive callback updates until the task reaches a terminal state.Kling 2.6 API for Video and Native Audio Workflows

Choose a valid Kling 2.6 request mode
Prompt-led text-to-video
https://api.goenhance.ai/api/v1/videos/generations with model: kling-2-6 and a required prompt of up to 2,500 characters. Omit image_url for text-to-video and select 16:9, 9:16, or 1:1 through ratio.
Image guidance, audio, and voices
image_url for image-to-video. Native audio requires 1080p, while a first-and-last-frame request forces 720p; those two modes cannot be combined. Up to two designated voices can be used with an input image and generate_audio: true.
Kling 2.6 API capabilities and validation rules
Model and prompt
model to kling-2-6. A prompt is required and accepts up to 2,500 characters.Input frames
image_url for a first frame. Optional image_end_url requires that first frame and forces 720p.Duration
Resolution
quality is accepted as an alias.Aspect ratio
Audio and voices
voice_ids require an input image and audio enabled; use @1 and @2 in the prompt.Kling 2.6 API pricing
720p without native audio
1080p without native audio
1080p with native audio
Integrate Kling 2.6 in six steps
1. Create an API key
2. Select the input mode
3. Validate constraints
4. Submit the task
/api/v1/videos/generations and persist the returned img_uuid.5. Track status
GET /api/v1/jobs/detail or provide a publicly accessible HTTPS custom_callback_url.6. Handle terminal states
Developer use cases for Kling 2.6
Campaign clip variants
Product image animation
Localized voice-led scenes
Kling 2.6 API FAQ
What endpoint does the Kling 2.6 API use?
Send a POST request to https://api.goenhance.ai/api/v1/videos/generations with model set to kling-2-6. A successful submission returns data.img_uuid.
What is the Kling 2.6 API pricing?
Without audio, pricing is 2.1 tokens/s ($0.042/s) at 720p and 3.5 tokens/s ($0.07/s) at 1080p. At 1080p, native audio is 7 tokens/s ($0.14/s), or 8.4 tokens/s ($0.168/s) with designated voices.
Can I combine a last frame with native audio?
No. A first-and-last-frame request forces 720p, while native audio requires 1080p. Remove image_end_url or disable audio.
How do designated voices work?
Supply up to two voice_ids with an input image and generate_audio: true. Their order maps to @1 and @2 in the prompt. A single voice_id alias is also accepted.
Which duration, resolution, and ratio values are supported?
Duration accepts 5 or 10 seconds. Resolution accepts 720p or 1080p. Text-to-video ratios are 16:9, 9:16, and 1:1; image-guided requests use the input image's ratio.
How do callbacks work?
Provide a public HTTPS callback URL. GoEnhance sends job-detail-shaped updates for processing, success, and failure; non-200 responses are retried up to three times with a three-second timeout per attempt.
Is this Kuaishou's official API?
No. This page describes independent access to Kling 2.6 through GoEnhance. GoEnhance is not presented as Kuaishou or Kling's official API, partner, or model developer.
