Kling Video 3.0
Key facts · verified Aug 15, 2026
Kling Video 3.0 is a Kuaishou video generation model released on 6 Feb 2026. Pricing: Free Basic tier; Standard $6.99/mo (660 credits); Pro $25.99/mo (3,000 credits); Premier $64.99/mo (8,000 credits); Ultra $127.99/mo (26,000 credits) - first-subscription monthly rates, annual billing cheaper. Source: kling.ai.
Kuaishou's Kling AI video-generation model (VIDEO 3.0 in the Kling 3.0 Model Series, released 6 Feb 2026), generating 3-15 second clips from text, image, or start/end-frame prompts with native synchronized audio in five languages and multi-shot storyboarding, plus native 4K output added 23 April 2026. Consumer plans run from a free Basic tier to Ultra at $127.99/mo (26,000 credits/mo, first-subscription rate).
Free Basic tier; Standard $6.99/mo (660 credits); Pro $25.99/mo (3,000 credits); Premier $64.99/mo (8,000 credits); Ultra $127.99/mo (26,000 credits) - first-subscription monthly rates, annual billing cheaper
Not published
Not rated
Not ranked
Key Features
Full Specifications
Free Basic tier; Standard $6.99/mo (660 credits); Pro $25.99/mo (3,000 credits); Premier $64.99/mo (8,000 credits); Ultra $127.99/mo (26,000 credits) - first-subscription monthly rates, annual billing cheaper | |
|---|---|
| Identity | |
| Developer | Kuaishou |
| Released | Feb 2026 |
| Status | GA - launched worldwide February 6, 2026 across Kling's consumer app and developer API; no beta/preview label found on the model guide, the API capability map, or the launch announcement |
| Licence | Proprietary |
| Self-hostable | No |
| Cost | |
| Price per second | API (pay-as-you-go, no subscription): No Native Audio - $0.084/s (720p), $0.112/s (1080p), $0.42/s (4K); With Native Audio (no Voice Control) - $0.126/s (720p), $0.168/s (1080p), $0.42/s (4K). The consumer app instead bills in subscription Credits rather than a flat $/s rate: No Native Audio - 6 Credits/s (720p), 8 Credits/s (1080p); Native Audio - 9 Credits/s (720p), 12 Credits/s (1080p); +2 Credits/s for optional Voice Control; a Credit costs $0.0049-$0.0106 depending on plan tier (see Cheapest plan) |
| Cheapest plan | Standard $6.99/mo (660 credits/mo; first-subscription rate, then $8.80/mo) is the cheapest paid consumer subscription; the API has no subscription tier - pure metered per-second billing (see Price per second) |
| Free tier | Yes, nominally - a Basic plan ($0/mo) exists, but Kling's own plan-comparison table shows no monthly credit allotment for it (shown as "-" under both "Daily Free Credits for Subscribers" and "Monthly Credits included", benefits that only start at Standard), and free-tier output is excluded from the "Generated content is for commercial use" benefit granted to every paid tier (see Commercial rights) |
| Capacity | |
| Max duration | 15 seconds per single generation (flexible duration, 3-15s); consumer subscription plans (Standard and above) separately list a "Video extension" benefit with no stated maximum extended length; Kling's own API capability map does not list Video Extension as supported for any current model, including Kling 3.0 |
| Max resolution | Up to native 4K (3840x2160) - Kling's own API capability map lists Kling 3.0 (kling-v3) as supporting 720p, 1080p, and 4K output; native 4K was added April 23, 2026 as a Kling 3.0 model-series upgrade, after the original February launch (whose own model-guide pricing table lists only 1080p/720p modes) |
| Capability | |
| Text-to-video | Yes - primary interface accepts natural-language text prompts; supports flexible 3-15s duration plus automatic "Multi-Shot" or manually-configured "Custom Multi-Shot" storyboarding across scenes |
| Image-to-video | Yes - image-to-video and start/end-frame-to-video, plus up to 3 bound "Element Reference" images/videos per generation for locked subject/character consistency through camera movement; Kling's API capability map separately confirms First/Last Frame and Element Control support for kling-v3 |
| Native audio | Yes - synchronized dialogue, sound effects and ambient audio generated natively alongside video, in 5 languages (Chinese, English, Japanese, Korean, Spanish) plus Chinese dialects and English accents, with multi-character coreference (3+) so each character's dialogue is matched to the right speaker |
| Camera controls | Prompt-based only - no dedicated camera-control parameter; Kling's own API capability map explicitly lists "Camera Control" as unsupported ("-") for every current model, including Kling 3.0, across both the Text-to-Video and Image-to-Video sections |
| Rights | |
| Commercial rights | Paid subscribers only - Kling's Terms of Use state that, without written permission, a user "may not use...the Output for any commercial purposes" by default (Section 4.6); the paid subscription tiers (Standard through Ultra) separately list "Generated content is for commercial use" as an included plan benefit, while that benefit is blank/absent for the free Basic tier in Kling's own plan-comparison table |
| Access | |
| API | Yes - Kling AI API Platform (kling.ai/dev), model code kling-v3, documented with a full capability map and per-mode/per-resolution per-second pricing |
| Chat app | Yes - Kling AI web app (kling.ai) plus native iOS, Android, macOS, and Windows apps |
| Cloud marketplaces | Not offered - not found on AWS Bedrock, Microsoft Foundry (Azure), or Google Cloud Vertex AI as of this verification |
| Fine-tuning | Not offered - no fine-tuning/custom-training API endpoint is documented; character/subject consistency instead comes via the separate Element Library reference-image/video system (create an Element once, reuse it via Element Reference - see Image-to-video), not a base-model fine-tune |