PARAS MADAN
All writing

Latest in AI

Kling 4.0 Is Fully Available: MCP, CLI and What Is Still Coming

Kling's wider rollout brings video generation into assistant and terminal workflows, while its enterprise API and HDR output remain on the roadmap.

Kling announced full availability of Kling 4.0 on October 15, 2026, alongside a refreshed creation interface and access through MCP and CLI. That gives creators ways to generate video from assistants and scripts. The same announcement still lists enterprise API access and 10-bit HDR output as coming soon.

The distinction matters if you are choosing how to build around it. A working connector can make a creative workflow easier today, while a product that depends on a dedicated enterprise API still needs to wait for that release's terms and documentation.

What changed in the Kling 4.0 rollout?

Kling 4.0 was introduced in early access on September 30. October 15's official announcement marks the wider rollout and highlights the new interface, MCP and CLI access. It should not be read as the model's first announcement.

The model's published capabilities include native video generation up to 30 seconds, up to 10 keyframes and prompts up to 8,000 tokens. Its reference system supports a combined maximum of 15 items, with separate limits for images, videos and reusable subjects. Those are creative controls, rather than a guarantee that every complex scene will render correctly.

For a product video, the practical benefit is being able to describe the sequence and supply visual references together. A longer shot reduces some stitching work, but the resulting clip still needs review for product details, motion and continuity.

Kling 4.0 availability at a glance

CapabilityStatus in the October 15 announcement
Kling 4.0Fully available, according to Kling
Refreshed creation interfaceAvailable
MCP and CLI accessAvailable
10-bit HDR output at 4K and 1080pComing soon
Enterprise and developer API accessComing soon

Full availability does not mean every announced feature has shipped. In particular, ordinary resolution support should not be confused with the forthcoming 10-bit HDR output. The announcement also does not establish a universal free entitlement or an enterprise API price.

How Kling MCP and CLI work

The official setup guide describes MCP as the connection through which assistants such as Claude and Cursor can invoke Kling's generation tools. CLI exposes those capabilities through terminal commands, which makes scripted and batch workflows possible.

Both routes use an authorized Kling account. They currently accept paid Personal-workspace credits, excluding Team benefits and bonus credits. Off-peak free generation is unavailable through assistants. The guide recommends choosing one route per assistant to avoid unpredictable tool selection.

Generation is asynchronous. A request returns a generation identifier, and a separate task query retrieves status and the eventual asset URL. The CLI also offers a polling option. A successful submission therefore tells you that a job was accepted, not that a usable video exists yet. Submitted jobs cannot be canceled, and generated file URLs expire after 24 hours, so completed assets need downloading promptly.

That changes the workflow design. A script producing several product clips should keep each job identifier, wait for completion and record the returned assets before deciding whether anything needs another attempt. Treating every slow response as a failed generation can create unnecessary work and spending.

Check model access before hardcoding a workflow

Kling's tool protocol examples make account-level capability discovery central. The who_am_i tool returns available models and their parameter specifications. Model identifiers and accepted inputs should come from that live response, rather than from an example copied into a script.

The documented generation response also includes credits consumed. Automation makes budget control more important, especially when an assistant can prepare many variations faster than a person can review them. Set a generation budget and agree on final settings before submitting a batch.

A useful first workflow is deliberately small: one approved reference image, one short product shot and a clear review criterion. Check whether the output preserves the product and communicates the intended action before increasing the number of variants. Cost per accepted clip will be more informative than the number of jobs an integration can submit.

Who should use the new access routes?

MCP is useful for creators who already plan work inside an assistant. CLI is more relevant when a team wants repeatable asset preparation and generation steps in an existing toolchain. Neither removes the need to inspect the finished media or understand the account's limits.

For developers evaluating a customer-facing integration, the next important release is the enterprise API documentation: authentication, quotas, pricing and operational commitments. For creators, the immediate test is simpler. Does bringing generation into the existing workflow reduce the time between a brief and an approved clip?

I’m Paras. I work at the intersection of AI, engineering and distribution.

I help technical founders turn launches and product updates into clear stories that developers and founders want to read and share.

Book a 30-minute call.