AlibabaAlibaba video model

Wan 2.6. Alibaba video with sound, from eight cents a second.

Wan 2.6 renders 5 or 10 second clips with native sound, in 16:9 or 9:16, from a prompt or a first frame.

  • 5 to 10 second clips
  • 16:9 and 9:16
  • Native audio
  • First frame control

Wan 2.6 is the Wan release before 2.7 and 3.0. It renders 5 or 10 second clips in 16:9 or 9:16, with ambience and effects generated alongside the picture, from a prompt or from a still you hand it as the first frame. OpenRouter prices it by resolution and input: 8 cents per second at 720p from a prompt and 10 from a still, 12 and 15 at 1080p.

In our test a 5 second vertical clip at 720p took 33 seconds, billed exactly 40 cents, and came back at 720 by 1280 with an audio track. On Treza it sits in the Video Generation node's picker beside Wan 2.7 and Wan 3.0, and you can ask for it by name in chat.

Sound with the picture

Ambience and effects are generated with the clip. Turn audio off per generation when a narrator or music bed will sit on top.

Vertical or wide

16:9 for YouTube and the web, 9:16 for Shorts, Reels and TikTok, from the same prompt.

Start from a still

Hand it an image as the first frame and it animates from there, so a product photo or a generated still keeps its composition.

Priced by what you ask for

8 cents per second at 720p from a prompt, 10 from a still; 12 and 15 at 1080p. Draft at 720p and step up only for the keeper.

What to make with Wan 2.6

Real prompts, and what comes back. Every one runs in Treza chat or as a node on the canvas.

“A barista pours steamed milk into a latte in a sunlit cafe, slow push-in, natural ambient sound, 9:16”

A vertical cafe clip with the room's sound.

“Animate this product still (attached): slow orbit, light sweeping across”

Image to video from your first frame.

“A night market in the rain, sizzling grills, people passing, 16:9”

Busy b-roll with generated street sound.

“A cat stretches on a sunny windowsill, birds chirping outside, 9:16”

A short vertical pet clip with ambience.

“Ocean waves crash against a lighthouse at dusk, 10 seconds”

A longer establishing shot with surf sound.

“The same lighthouse shot on Wan 3.0, 20 seconds”

A longer take of the same idea on the newer model.

Wan 2.6 is one node on a canvas

Chain it with language, image, and video models, then publish the whole pipeline as an API. See how a full AI video pipeline fits together.

  1. Step 01

    Draft the brief

    A language model node expands a one-line idea into a detailed, on-brand prompt before it ever reaches the model.

  2. Step 02

    Generate with Wan 2.6

    Run variants in parallel, compare against other models on the same prompt, and keep every result in run history.

  3. Step 03

    Publish it as an API

    The pipeline becomes a versioned endpoint. Your product calls it with the typed /invoke API or any OpenAI SDK.

Wan 2.6 as an API, priced per clip

The Wan 2.6 API

Build a pipeline with Wan 2.6 on a node and publish it. It becomes a versioned HTTP endpoint: the typed /invoke route enqueues a run and returns a run id to poll, and an OpenAI-compatible route works with SDKs you already use. Agents connect over MCP the same way. Your prompt engineering ships inside the pipeline, so every caller gets your quality, and swapping the model later does not change the API your product calls.

Wan 2.6 pricing

Prepaid credits, no subscription. Each clip is charged at Wan 2.6's own rate from your balance, shown before you run, and a failed generation costs nothing. Credit packs start at $5 and never expire, which suits spiky production schedules. Run history records the cost of every generation, per node, so spend is auditable rather than a surprise.

Wan 2.6, answered

What does Wan 2.6 cost per second?

OpenRouter lists 8 cents per second at 720p from a prompt and 10 cents from a first frame, and 12 and 15 cents at 1080p. A 5 second 720p clip from a prompt is 40 cents, which is what our test billed. On Treza it is charged from prepaid credits, and a failed generation costs nothing.

How is Wan 2.6 different from Wan 2.7 and Wan 3.0?

Wan 2.6 renders 5 or 10 seconds in 16:9 or 9:16 from a prompt or a first frame. Wan 2.7 adds reference images, including photos of real people, and a last frame. Wan 3.0 holds a single take for up to 30 seconds in five aspect ratios. All three are in the same picker and in chat.

Can I use Wan 2.6 in chat?

Yes. Ask for Wan 2.6 by name in Treza chat. On the canvas it is a model in the Video Generation node, and a pipeline built on it can be published as an API endpoint.

How much does Wan 2.6 cost on Treza?

Treza uses prepaid credits with no subscription. Each clip is charged at Wan 2.6's rate from your balance, and only successful generations are charged. Credit packs start at $5 and never expire.

Can I call Wan 2.6 through an API?

Yes. Build a pipeline that uses Wan 2.6 and publish it as a versioned HTTP endpoint. Call the typed /invoke API from your product, or point any OpenAI SDK at the OpenAI-compatible endpoint. If a better model ships later, swap it on the node without changing the API your product calls.