# Gemini Omni 1.2 Flash API: access, pricing and code

> Get the official Gemini Omni 1.2 Flash API from Google on Runway Dev: model ID, endpoints, price per second by resolution, inputs and code.

Author: Runway · Published: 2026-11-09 · Updated: 2026-11-09 · Source: https://dev.runway.com/learn/gemini-omni-flash-1-2-api

## Summary

Runway Dev gives you the official Gemini Omni 1.2 Flash API from Google, available since the day its API launched.\
\
Gemini Omni 1.2 Flash is Google's multimodal video model: it generates clips with native audio from a text prompt, from a first and a last frame or from image and video references, and it extends or edits clips you already have.\
\
This page covers access, pricing, inputs and code, and how your coding agent can build the integration for you.

## What is Gemini Omni 1.2 Flash?

Gemini Omni 1.2 Flash is Google's video generation and editing model, and the successor to [Gemini Omni 1.1 Flash](https://dev.runway.com/learn/gemini-omni-flash-1-1-api). It reads text, images and video together and returns a clip of 3 to 15 seconds with audio generated in the same pass: dialogue written into the prompt, sound effects and the sound of the scene.

![A slow river of glowing lava creeps over black volcanic rock at dusk: its dark crust cracks open and a chunk of cooled rock breaks off the front and tumbles down.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/gemini-omni-flash-1-2-api-hero-lava-0c94e3d6.mp4 'One text-to-video request: the crackle, the splitting crust and the falling rock, generated with the picture.')

```text
One continuous locked-off close shot, with no cuts, of a slow river of glowing lava creeping over black volcanic rock at dusk. The dark crust cracks and splits open to show the bright orange lava inside, and a chunk of cooled rock breaks off the front and tumbles down. No people. Audio: a low crackling and hissing, sharp cracks as the crust splits, the clatter of the falling rock.
```

Google [made it generally available in November 2026](https://cloud.google.com/blog/products/ai-machine-learning/gemini-omni-1-2-flash-generally-available), and Runway Dev offered it the same day. It follows image and video references more closely than 1.1, so a product or a character stays the same from shot to shot, and its clips run longer than 1.1's. It generates the motion between a first and a last frame, continues a clip you already have and edits one from a text instruction. Output comes at 360p, 720p, 1080p and 4K, the smallest being a low-cost size for drafts.

## What are the ways to use Gemini Omni 1.2 Flash on Runway Dev?

Gemini Omni 1.2 Flash runs on Runway Dev four ways, all from one account:

| Way in                                                                                                                 | Best for                                                                                          | What you need                                                |
| ---------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | ------------------------------------------------------------ |
| [Playground](https://dev.runway.com/endpoints/text_to_video?modelId=gemini_omni_flash_1.2)                             | Trying a prompt, then copying the request as code                                                 | A Runway Dev account                                         |
| [API](https://docs.dev.runwayml.com/guides/using-the-api/) and [SDKs](https://docs.dev.runwayml.com/api-details/sdks/) | Calling Gemini Omni 1.2 Flash from your own code                                                  | An API key, plus the Node or Python SDK if you want one      |
| [Coding agent](https://dev.runway.com/learn/build-with-your-coding-agent)                                                                    | Having Claude Code, Codex or Cursor build the API integration into your product, tool or pipeline | The Runway Dev MCP and skills, and an API key for live calls |
| [Workflows](https://dev.runway.com/workflows)                                                                                                | Running Gemini Omni 1.2 Flash as one step of a multi-model pipeline, with a single API call       | A workflow built in the Runway app and published             |

Gemini Omni 1.2 Flash is also available in [Model Router](https://dev.runway.com/model-routers), a saved config you send with the request instead of a model ID. You set what to optimize for (cost, quality or latency), a maximum cost per generation and the models allowed, and the router picks one for each request sent to `POST /v1/generate/video`.

## How do you get API access to Gemini Omni 1.2 Flash?

You get API access to Gemini Omni 1.2 Flash with a Runway Dev account and an API key:

1. Create an account at https://dev.runway.com.
2. Create an API key in your project's dashboard.
3. Add credits to your project's balance.
4. Store the key in your environment as `RUNWAYML_API_SECRET`.
5. Send a request with `"model": "gemini_omni_flash_1.2"` and the `X-Runway-Version` header to one of its endpoints. The code below is ready to run.

## How much does the Gemini Omni 1.2 Flash API cost?

The Gemini Omni 1.2 Flash API on Runway Dev is billed per second of the video it makes, at a rate set by the output size: 10 credits ($0.10) per second at 1080p, with lower rates at the smaller sizes and a higher one at the largest.

| Resolution | Credits | USD |
| --- | --- | --- |
| 360p | 2 credits per second | $0.02 per second |
| 720p | 7 credits per second | $0.07 per second |
| 1080p | 10 credits per second | $0.10 per second |
| 4K | 21 credits per second | $0.21 per second |

- Reference images add 1 credit ($0.01) each; input video adds 1 credit ($0.01) per second.

Prices current as of November 10, 2026. 1 credit = $0.01.

![Red and yellow autumn leaves float on a slow, clear forest stream and drift past a mossy rock.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/gemini-omni-flash-1-2-api-price-leaves-4k-9a50714c.mp4 'Gemini Omni 1.2 Flash at its largest size: 21 credits ($0.21) per second at 4K.')

```text
Locked-off close shot of red and yellow autumn leaves floating on a slow, clear forest stream, carried by the current past a mossy rock, fine ripples catching the light. Audio: gentle running water, a few birds.
```

How the billing works:

- **Sound comes with every clip at the same rate.** Audio is generated each time, and there's no audio option or separate audio price.
- **Images add a charge per image**, references and keyframes alike: on image-to-video, the first frame and the last frame each count as one.
- **Input video is billed per second**, on top of the output: the clip you extend, edit or use as a reference.
- **A `duration` of `auto` is billed as 15 seconds up front**, and what the clip doesn't use is refunded once it finishes. An edit is billed on the length of its source clip instead.
- **An extension is billed for the footage it adds.** The file it returns holds your clip followed by the new footage, and `duration` sets only the new part.
- **Extend and edit are billed at the size you ask for** in `resolution`, 720p when you leave it out.
- **A task that fails input moderation (`SAFETY.INPUT.*`) is still charged.** See [task failures](https://docs.dev.runwayml.com/errors/task-failures) for what each failure means.

Every rate is also listed on [dev.runway.com/pricing](https://dev.runway.com/pricing) and in the [docs pricing guide](https://docs.dev.runwayml.com/guides/pricing/).

## What inputs does Gemini Omni 1.2 Flash accept?

Gemini Omni 1.2 Flash needs a prompt of up to 4,000 characters on every endpoint and takes a duration of 3 to 15 seconds or `auto`. Images and clips go in as frames, as references or as the video to work on, and each endpoint uses them its own way.

On `POST /v1/text_to_video` the prompt describes the picture and the sound together. A request can add up to 5 images and 3 videos as references, images in `references` and clips in `referenceVideos`, to set who or what appears and how it looks. Each reference clip runs up to 10 seconds.

On `POST /v1/image_to_video` you set the opening frame, and the closing one if you want it. A single `promptImage` URL is the first frame. To set both, send a list with one image whose `position` is `first` and one whose `position` is `last`, and Gemini Omni 1.2 Flash generates the motion between them. A last frame on its own is rejected.

![A rocky shore at low tide in late afternoon, seaweed on dark rocks and small tide pools reflecting the sky.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/images/gemini-omni-flash-1-2-api/tide-first-afbd8669.jpg 'First frame') ![The same rocky shore at high tide, the sea covering the lower rocks and the tide pools.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/images/gemini-omni-flash-1-2-api/tide-last-d6a53902.jpg 'Last frame')

![The tide comes in over a rocky shore in one fixed shot, filling the tide pools and covering the lower rocks.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/gemini-omni-flash-1-2-api-keyframes-tide-bdac42ba.mp4 'Gemini Omni 1.2 Flash generates the change between the first and the last frame.')

```text
Time-lapse from a fixed viewpoint as the tide comes in over the rocky shore: each set of waves reaches further, fills the tide pools and covers the lower rocks until only the tallest stand above the water. Audio: waves breaking and washing over the rocks, growing louder.
```

On `POST /v1/video_to_video`, `mode` decides what happens to the clip in `videoUri`. `reference` (the default) makes a new video guided by it, and takes a `ratio` and image references. `extend` continues the clip, with `duration` setting how much footage it adds. `edit` changes the clip in place from the prompt and returns a video as long as the source, so its `duration` must be `auto`. Extend and edit keep the input's orientation, take a `resolution` in place of `ratio` and render at any of the model's sizes, and extend takes no image references.

![A gray heron stands in the shallows of a misty lake at dawn, then spreads its wings and flies low across the water and out of frame.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/gemini-omni-flash-1-2-api-extend-heron-165721a4.mp4 'One extend request: the first six seconds are the clip it was sent, the last six are the takeoff and the sound it added.')

```text
The heron spreads its wings, lifts off from the water and flies low across the lake and out of frame to the right, ripples spreading where it stood. Audio: the beat of its wings, splashing water, then quiet.
```

The length of the input clip depends on the mode: `reference` takes a clip of up to 10 seconds, `edit` takes a clip of under 10 seconds, and `extend` takes a clip of up to 30 seconds, so an extended clip can be extended again while it stays under that length.

Here is every field, by endpoint:

### `POST /v1/text_to_video`

| Field | Accepts |
| --- | --- |
| `model` (required) | `gemini_omni_flash_1.2` |
| `promptText` (required) | Text, 1 to 4,000 characters |
| `ratio` | One of `640:360`, `360:640`, `1280:720`, `720:1280`, `1920:1080`, `1080:1920`, `3840:2160` or `2160:3840` |
| `duration` | Whole number from 3 to 15, or `auto` |
| `references` | A list of up to 5 images |
| `referenceVideos` | A list of up to 3 videos |

### `POST /v1/image_to_video`

| Field | Accepts |
| --- | --- |
| `model` (required) | `gemini_omni_flash_1.2` |
| `promptImage` (required) | One image, or a list of up to 2 images, each with a `position` of `first` or `last` (a `last` frame needs a `first` one) |
| `promptText` (required) | Text, 1 to 4,000 characters |
| `ratio` | One of `640:360`, `360:640`, `1280:720`, `720:1280`, `1920:1080`, `1080:1920`, `3840:2160` or `2160:3840` |
| `duration` | Whole number from 3 to 15, or `auto` |

### `POST /v1/video_to_video`

| Field | Accepts |
| --- | --- |
| `model` (required) | `gemini_omni_flash_1.2` |
| `videoUri` (required) | One video |
| `promptText` (required) | Text, 1 to 4,000 characters |
| `mode` | One of `reference`, `extend` or `edit`. Default `reference` |
| `resolution` | One of `360p`, `720p`, `1080p` or `4k` |
| `ratio` | One of `640:360`, `360:640`, `1280:720`, `720:1280`, `1920:1080`, `1080:1920`, `3840:2160` or `2160:3840` |
| `duration` | Whole number from 3 to 15, or `auto` |
| `references` | A list of up to 5 images |

Images and videos go in as an HTTPS URL, a `runway://` URI from an [upload](https://docs.dev.runwayml.com/assets/uploads) or a data URI; [inputs](https://docs.dev.runwayml.com/assets/inputs) lists sizes and formats. Each table shows what a field accepts on its own. Rules that combine fields are checked when the task is submitted, and the [API reference](https://docs.dev.runwayml.com/api/) has them in full.

A few rules change what you send:

- **`ratio` sets the size as well as the shape**, and the size sets the price. Each value is an output size in pixels, landscape or portrait, and a request without one renders at 720p landscape.
- **`duration` defaults to `auto`**, so the model picks the length unless you send a number.
- **There's no `audio` field.** Every clip has sound, and a request that sends `audio` returns a 400.
- **Unknown fields return a 400.** The request schema is strict, so a field name from another API, such as `aspect_ratio`, is rejected rather than ignored.
- **Frames must fit the aspect range in [inputs](https://docs.dev.runwayml.com/assets/inputs).** A first or last frame far wider or taller than that is rejected.

## How do you use the Gemini Omni 1.2 Flash API?

You use the Gemini Omni 1.2 Flash API by sending `gemini_omni_flash_1.2` as the model to one of its endpoints. This example request, in curl, Node and Python, makes a clip with sound from text with `POST /v1/text_to_video`, and the other fields keep their defaults:

**curl**

```bash
curl -X POST https://api.dev.runwayml.com/v1/text_to_video \
  -H "Authorization: Bearer $RUNWAYML_API_SECRET" \
  -H "X-Runway-Version: 2024-11-06" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini_omni_flash_1.2",
    "promptText": "One continuous locked-off wide shot, with no cuts, of a glacier face above a still fjord on a gray morning. A large slab of ice breaks away from the face and falls into the water, and a wave rolls out across the fjord toward the rocky shore in the foreground. Audio: a sharp crack, the roar of the ice hitting the water, the wave washing against the rocks."
  }'

# The response carries the task id. Poll the task until it finishes:
curl https://api.dev.runwayml.com/v1/tasks/$TASK_ID \
  -H "Authorization: Bearer $RUNWAYML_API_SECRET" \
  -H "X-Runway-Version: 2024-11-06"
```

**Node.js**

```javascript
import RunwayML, { TaskFailedError } from '@runwayml/sdk';

// Reads your API key from RUNWAYML_API_SECRET.
const client = new RunwayML();

try {
  const task = await client.textToVideo
    .create({
      model: 'gemini_omni_flash_1.2',
      promptText: 'One continuous locked-off wide shot, with no cuts, of a glacier face above a still fjord on a gray morning. A large slab of ice breaks away from the face and falls into the water, and a wave rolls out across the fjord toward the rocky shore in the foreground. Audio: a sharp crack, the roar of the ice hitting the water, the wave washing against the rocks.',
    })
    .waitForTaskOutput();
  console.log(task.output);
} catch (error) {
  if (error instanceof TaskFailedError) {
    console.error(error.taskDetails);
  } else {
    throw error;
  }
}
```

**Python**

```python
from runwayml import RunwayML, TaskFailedError

# Reads your API key from RUNWAYML_API_SECRET.
client = RunwayML()

try:
    task = client.text_to_video.create(
        model="gemini_omni_flash_1.2",
        prompt_text="One continuous locked-off wide shot, with no cuts, of a glacier face above a still fjord on a gray morning. A large slab of ice breaks away from the face and falls into the water, and a wave rolls out across the fjord toward the rocky shore in the foreground. Audio: a sharp crack, the roar of the ice hitting the water, the wave washing against the rocks.",
    ).wait_for_task_output()
    print(task.output)
except TaskFailedError as error:
    print(error.task_details)
```

![A slab of ice breaks from a glacier face, falls into a still fjord and sends a wave rolling toward the rocky shore.](https://d3phaj0sisr2ct.cloudfront.net/devportal/learn/gemini-omni-flash-1-2-api-code-glacier-a7ef35ae.mp4 'The request above, as Gemini Omni 1.2 Flash returns it.')

Generation is asynchronous:

- **The create request returns a task ID right away**, and the video is made in the background.
- **Poll `GET /v1/tasks/{id}` with backoff** until the status is `SUCCEEDED`, `FAILED` or `CANCELLED`. `THROTTLED` means the task is waiting in a queue, not rejected.
- **Output URLs are temporary**, so download the video when the task succeeds. See [outputs](https://docs.dev.runwayml.com/assets/outputs).
- **The SDKs poll for you.** `waitForTaskOutput()` and `wait_for_task_output()` raise `TaskFailedError` with the task's details when it fails. They also stop waiting after a default timeout while the task keeps running, so pass a longer timeout or keep polling the task ID.
- **To use your own rules instead of one fixed model**, send a [Model Router](https://dev.runway.com/learn/choosing-the-right-model)'s `configId` to `POST /v1/generate/video` in place of the model. Add `dryRun` to see which model it picks without creating a task.

## Why does a Gemini Omni 1.2 Flash API request fail?

A Gemini Omni 1.2 Flash request fails with a 400 when a field is wrong, and a task fails with a failure code that names the cause. The common ones:

- **A 400 that names `audio` or `aspect_ratio`**: the request carries a field Gemini Omni 1.2 Flash doesn't take on Runway Dev. Drop `audio`, since every clip has sound, and set the size and shape with `ratio`.
- **A 400 for `resolution` in reference mode, or for `ratio` in extend or edit mode**: reference mode sets the output with `ratio`, and extend and edit with `resolution`, since they keep the input's orientation.
- **A 400 saying the input video is too long for its mode**: `reference`, `extend` and `edit` each cap the input clip. Trim it to the mode's cap in the inputs above.
- **A 400 for image references in extend mode**: an extension takes none. Describe the new footage in the prompt instead.
- **A 400 for a number in `duration` in edit mode**: an edit returns a clip as long as the source. Send `auto`.
- **An image-to-video request with only a last frame**: a first frame is required. Add one with `position` set to `first`.
- **`SAFETY.INPUT.*`**: content moderation rejected an input. The task is charged and shouldn't be retried as is: change the inputs first.
- **`SAFETY.OUTPUT.*`**: the generated video was rejected by moderation. Change the prompt or references before you try again.
- **An input file that couldn't be decoded**: re-export it in a common format, such as an H.264 MP4 or a JPEG, and try again.
- **`THIRD_PARTY.UNAVAILABLE` or `THIRD_PARTY.RATE_LIMITED`**: the model's provider didn't return the video, because of an outage or load. Wait, then retry.
- **An output link that stopped working**: output URLs expire. Download the video when the task succeeds.

## What does Runway Dev add when you call Gemini Omni 1.2 Flash?

Runway Dev serves Gemini Omni 1.2 Flash next to [Runway's own models and other labs' models](https://dev.runway.com/models/catalog), behind one API key and one credit balance: you build with a single key and pay a single bill, instead of an account, a key and an invoice per lab. It has been on Runway Dev since the day Google's API for it launched.

A coding agent can do the Gemini Omni 1.2 Flash integration for you on Runway Dev, end to end. Connect the [Runway Dev MCP](https://docs.dev.runwayml.com/guides/mcp/) and the agent works from your live account: the models your project can call with their current prices and input limits, your credit balance and the record behind any task. It writes the integration against what the API accepts today, and when a task fails it looks up the reason and fixes it. The [Runway Dev skills](https://github.com/runwayml/skills) show the agent how to integrate with the official SDKs and how to build with models, Model Routers, Recipes and Workflows.

[Workflows](https://dev.runway.com/workflows) put Gemini Omni 1.2 Flash in a pipeline with Runway's own models and other labs' models, run with a single API call. A creative or marketing teammate builds the workflow visually in the [Runway app](https://app.runwayml.com/workflows) ([Introduction to Workflows](https://help.runwayml.com/hc/en-us/articles/45763528999699-Introduction-to-Workflows)), then [publishes it as an endpoint](https://help.runwayml.com/hc/en-us/articles/50682960972947-Publishing-a-Workflow-as-an-Endpoint) your code calls.

## FAQ

**What do I need to call the Gemini Omni 1.2 Flash API?**

Create a Runway Dev account at dev.runway.com, make an API key and add credits. Then send `gemini_omni_flash_1.2` as the model to any of its endpoints: `POST /v1/text_to_video`, `POST /v1/image_to_video` and `POST /v1/video_to_video`. To have a coding agent write the integration instead, connect the Runway Dev MCP and install the Runway Dev skills: the agent builds it from your live account, with the prices and input limits your project has today.

**Can I use the Gemini Omni 1.2 Flash API without a Google Cloud account?**

You can use the Gemini Omni 1.2 Flash API without a Google Cloud account: a Runway Dev account and API key are all you need. There's no Google Cloud project, billing account or region to set up, and Gemini Omni 1.2 Flash is paid from the same credit balance as every other model on Runway Dev.

**What does Gemini Omni 1.2 Flash cost per second at each resolution?**

On Runway Dev, Gemini Omni 1.2 Flash costs 21 credits ($0.21) per second at 4K, 10 credits ($0.10) per second at 1080p, 7 credits ($0.07) per second at 720p and 2 credits ($0.02) per second at 360p, with sound included at every size. Each image adds a per-image charge and input video a per-second one, and every generation rounds up to a whole credit. Prices are current as of November 10, 2026.

**Is the Gemini Omni 1.2 Flash API on Runway Dev the official model?**

Runway Dev's Gemini Omni 1.2 Flash API runs the official Gemini Omni 1.2 Flash model from Google. You call it with your Runway Dev API key and the model ID `gemini_omni_flash_1.2`, like every other model on Runway Dev.

**How is Gemini Omni 1.2 Flash different from Gemini Omni 1.1 Flash?**

Gemini Omni 1.2 Flash makes longer clips than [Gemini Omni 1.1 Flash](https://dev.runway.com/learn/gemini-omni-flash-1-1-api), follows image and video references more closely and costs less per second at every size on Runway Dev. Its extend and edit modes also render at any of its sizes. Both run on Runway Dev with the same API key: switching is the `model` field.

**Can I use Gemini Omni 1.2 Flash without writing code?**

You can use Gemini Omni 1.2 Flash on Runway Dev without writing code yourself. The playground sends a request from a form and gives it back as code. A coding agent connected to the Runway Dev MCP writes the integration for you, and a teammate can build a workflow around it in the Runway app.

**Does the same Runway Dev API key work for Gemini Omni 1.2 Flash and other models?**

One Runway Dev API key calls Gemini Omni 1.2 Flash and every other model your project can use, Runway's own and other labs', with the same SDKs, task flow and credit balance. Switching models means changing the `model` field, or the rules of a Model Router.

**What is the Gemini Omni 1.2 Flash model ID on Runway Dev?**

The model ID is `gemini_omni_flash_1.2`, and it is the same on all of its endpoints: `POST /v1/text_to_video`, `POST /v1/image_to_video` and `POST /v1/video_to_video`.

**How long can a Gemini Omni 1.2 Flash video be, and at what resolutions?**

A Gemini Omni 1.2 Flash clip on Runway Dev runs 3 to 15 seconds, or set `duration` to `auto` and the model picks the length. It comes out at 360p, 720p, 1080p and 4K, landscape or portrait, chosen through the `ratio` you send. To go longer, extend it on `POST /v1/video_to_video`: `extend` takes a clip of up to 30 seconds and adds up to 15 seconds to it.

**Can Gemini Omni 1.2 Flash extend or edit an existing video through the API?**

Gemini Omni 1.2 Flash extends and edits clips on `POST /v1/video_to_video`. Set `mode` to `extend` to continue a clip, with `duration` as the length to add, or to `edit` to change it in place from `promptText`. Both keep the input's orientation, and `resolution` sets their output size.

**Why did my Gemini Omni 1.2 Flash request fail with `SAFETY.INPUT.TEXT`?**

`SAFETY.INPUT.TEXT` means content moderation rejected an input of your request; the suffix hints at the prompt but may not match. The task is charged and shouldn't be retried as is, so change the prompt or inputs before you send it again.

**Can I switch from Gemini Omni 1.2 Flash to another video model without changing my code?**

A Model Router lets you change models without touching your code. Send its `configId` to `POST /v1/generate/video`, and change the models it may use, its priority or its cost cap in the router.