Access Gemini Omni Flash - also searched as Gemini Omni or Omni Flash - through EvoLink's unified video API. Test text-to-video, image-to-video, reference-to-video, and conversational video editing before you integrate.
Pick a mode, write a prompt, and see the estimated cost before you spend anything.
Create task
Fixed to gemini-omni-flash-text-to-videoAuto lets the provider pick the length; the estimate above assumes 10 seconds.
Results are kept for 24 hours — download anything you want to keep.
Gemini Omni Flash is billed by tokens, not by the second. The final charge always uses the tokens the model actually returned.
One 3s 720p sample
About 35 tests with $10 in credits.
In Playground, pick Text to Video, set Duration to 3s, and send a short prompt with no input media - that is the cheapest valid request on this page. Output is always 720p, so cost scales with duration, not resolution.
Validate a prompt and confirm the billing shape
Iterate a full storyboard across all four routes
Batch production for a short-form campaign
Estimates only. Gemini Omni Flash bills by token, so the final charge always uses the usage the model actually returned. Auto duration is reserved as 10s and settled against the seconds actually produced.
Gemini Omni Flash bills by token across three categories: video output, input (prompt + media), and other output (thinking tokens). Final charges always reflect the usage actually returned by the model.
| Item | Rule | Rate | Billing |
|---|---|---|---|
| Video Outputper 1K tokens | Generated video frames and audio | $0.015 / 1K tokens-15% 1.0115 cr / 1K tokens$0.018 Official | Output |
| Inputper 1K tokens | Prompt text, reference images, and input video | $0.0013 / 1K tokens-15% 0.0867 cr / 1K tokens$0.0015 Official | Input |
| Other Outputper 1K tokens | Thinking tokens and internal reasoning | $0.0077 / 1K tokens-15% 0.5202 cr / 1K tokens$0.0090 Official | Output |
Rates shown are per 1,000 tokens. The actual charge for each request is calculated from the usage values returned in the API response.
Use Gemini Omni Flash on EvoLink for text-to-video, image-to-video, reference-to-video, and video editing through one unified video API. Public discussion often frames Gemini Omni as a video counterpart to Nano Banana because it brings multimodal video creation and conversational editing into short-form workflows. On EvoLink, the practical value is API access: EvoLink model IDs, async task workflow, callback support, token-based usage visibility, and the same API key used for Veo, Seedance, Kling, and other video models.
gemini-omni-flash-text-to-videoGenerate a 720p video from a prompt with 3-10 second or Auto duration.
gemini-omni-flash-image-to-videoAnimate one input image with a prompt and 3-10 second or Auto duration.
gemini-omni-flash-reference-to-videoUse 1-6 reference images to guide a 3-10 second or Auto-duration video.
gemini-omni-flash-video-editEdit one MP4 input video with a natural-language instruction.
Generate a clip with Gemini Omni, then refine it in conversation — "make the lighting warmer", "replace the red car". The workflow is designed for iterative edits while preserving the surrounding scene, subject identity, and motion as much as the selected route supports.
Swap an object in frame, remove an unwanted element, or rewrite a scene while preserving identity and motion. Useful for ad creative iteration and product variant rendering without external editing tools.
Pass a reference image and Gemini Omni anchors character identity, lighting, and color across the generated video. Combine with chat-based editing to refine specific shots without losing visual consistency.
Gemini Omni Flash routes can return short video outputs with audio where supported by the selected mode, reducing the need to stitch a separate TTS or sound-design pipeline into first-pass generation.
Use the EvoLink API for product integration and batch jobs, or call it from Codex, Claude, or Gemini for fast creative and development workflows. Both paths share the same EvoLink API key, balance, model routes, and task history.
Best for: product backends, batch jobs, automated workflows
Call EvoLink’s unified video API from your server and control the model ID, parameters, task queue, callbacks, and result storage.
Best for: creative and development tasks in Codex, Claude, and Gemini
Give the Agent your output goal, assets, and acceptance criteria. It can choose the route, assemble the request, track the task, and return the result without requiring you to hand-code every step.
Four workflows that map onto the four routes. Each one is a different input shape, not a different quality tier - pick by what you have on hand.
Generate 3-10 second cuts from the prompt alone, then iterate on wording rather than re-shooting. Native audio means the first pass is already reviewable without a separate TTS or sound pass.
Image-to-video takes exactly one input image and adds controlled motion and lighting while holding the subject. Useful when you have approved product photography but no video budget.
Reference-to-video accepts 1-6 images to anchor a character, palette, or style across multiple generated shots. Order carries no meaning here - the set is read as a whole.
Video edit rewrites objects, style, or lighting in one MP4 up to 10 seconds. Length and aspect ratio follow the input, so the output stays drop-in compatible with the original cut.
This example shows the shortest runnable flow: create a task, save the returned task ID, then poll or wait for an HTTPS callback. Open the API tab for the complete parameter and response reference.
curl -X POST https://api.evolink.ai/v1/videos/generations \
-H "Authorization: Bearer $EVOLINK_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gemini-omni-flash-text-to-video",
"prompt": "A cinematic wide shot of a glass greenhouse at sunrise, slow camera push in",
"duration": 8,
"aspect_ratio": "16:9",
"callback_url": "https://your-domain.com/webhooks/video"
}'
# Save the returned task ID, then query:
curl https://api.evolink.ai/v1/tasks/{task_id} \
-H "Authorization: Bearer $EVOLINK_API_KEY"Check every value against the parameter cards above: ranges, allowed options, asset counts, and file types must match the selected route.
Check the Authorization bearer token and confirm the available balance and task billing in the console.
Review likeness rights, brand assets, readable text, sensitive content, and input-asset compliance.
A Gemini Omni Flash task that reaches a final failed state is not charged. Keep the task ID, inspect the final state, and identify whether the cause is validation, review, or execution before retrying.
callback_url must be a public HTTPS URL, not localhost or a private network address. Always keep the task ID as a polling fallback.
Three models commonly shortlisted for production video workflows in 2026. All three accessible through one EvoLink API key.
| Feature | Gemini Omni | Veo 3.1 | Seedance 2.0 |
|---|---|---|---|
| EvoLink price | Token-based | From $0.04/s | From $0.045/s |
| Quality | 720p | 720p / 1080p, 4K upscaling where available | 480p / 720p / 1080p |
| Native audio | Yes | Yes | Yes |
| Reference control | Text + image + chat edit | Text + image | Text + image + video + audio |
| Video length | 3-10s / Auto | Short clips with Extend for longer scenes where supported | 4–15s |
| Editing | Conversational editing workflow | Generation-first | V2V mode |
| Best for | Short-form editing and multi-input workflows | Cinematic baseline | Multimodal reference production |
Three steps to your first Gemini Omni video task. Same integration pattern as Veo 3.1, Seedance 2.0, and Kling 3.0.
Sign up on EvoLink.ai and generate your API key from the dashboard. No Google Cloud project required.
POST to /v1/videos/generations with one of the Gemini Omni Flash model names and your prompt. Add duration for 3-10 second or Auto generation modes, image_urls for image-to-video or reference-to-video, video_urls for video edit, and callback_url for completion notification. The API processes asynchronously and returns a task id.
Use the task ID to poll the status endpoint, or wait for the callback_url webhook. When status reaches completed, you receive a download URL for the generated MP4. Links are valid for 24 hours.
Technical specifications for production video workflows.
Editing
Multi-turn refinement in a conversational workflow, with scene continuity depending on the selected route and input quality.
Output
720p output with configurable 3-10 second or Auto clips for generation modes. Auto is estimated as 10 seconds. Video edit accepts one MP4 input up to 10 seconds.
Modes
T2V from prompts and I2V with reference image input. Chat editing applies to outputs of either mode.
Audio
Short video outputs can include audio where supported by the selected Gemini Omni Flash route.
Consistency
Designed for stronger continuity across multi-input and edit-heavy workflows; validate consistency on your own production prompts.
Workflow
Submit a task, receive an ID, poll status or configure a callback_url. Same lifecycle as other EvoLink video models.
100 × 3-10s/Auto clips for social media batch
Use current Pricing tab rates
1,000 × 3-10s/Auto clips/month at production scale
Use current Pricing tab rates
1 generation + 3 edits multi-turn workflow
Use current Pricing tab rates
Use the Pricing tab above for current token-based rates. Select the workflow by changing the model parameter.
Gemini Omni is most interesting for workflow rather than raw fidelity alone: multimodal inputs, conversational editing, and a practical EvoLink route for testing it beside Veo, Seedance, and Kling with one API key.
Gemini Omni is positioned around conversational video editing, while Veo 3.1 and Seedance 2.0 are usually evaluated first as generation routes. For multi-turn refinement, this is the workflow difference to test.
Gemini Omni is reported to benefit from Gemini context and world knowledge for continuity across multi-input and edit-heavy workflows. Treat this as a behavior to evaluate in your own storyboard or short-video pipeline.
No GCP setup, no Vertex billing, no separate region approval. If you already run video generation through EvoLink, adding Gemini Omni is a one-parameter change — same request shape, same task lifecycle as Veo 3.1, Seedance 2.0, and Kling.
Compare Google video routes under one EvoLink account and API key, then choose by shot duration, resolution, and cost.
Gemini Omni Flash
Current modelThis page. Four routes at 720p with native audio, 3-10s or Auto duration, and conversational video editing.
Gemini Omni 1.1 Flash
The newer Omni generation: 360p to upscaled 4K, first-and-last-frame control, plus video edit and extend.
View API
Veo 3.1
Google's cinematic video model with native audio, for longer-form and higher-fidelity shots.
View APIWhen Gemini Omni Flash is not the default, use the same brief to compare resolution, shot duration, audio, motion stability, and total project cost.
Seedance 2.5
For 4-30 second shots with up to 50 multimodal references; compare usable-shot rate on longer clips.
View API
MiniMax H3
For detailed 2K clips of 4-15 seconds; compare its usable-shot rate with the same brief.
View API
Kling 3.0
For creative generation, action, and multi-shot video; a production fallback outside the Gemini family.
View API
Topaz Video Upscale
When a shot already has usable motion and composition but looks soft, use Topaz for detail recovery and 1x, 2x, or 4x upscaling.
View APIGemini Omni is Google's multimodal video model family announced at Google I/O 2026, with Omni Flash discussed as a short-form video route for text, image, video, and audio inputs. Compared with Veo 3.1, Gemini Omni is more interesting for conversational editing and multi-input workflows, while Veo remains a strong cinematic generation baseline.
Billing follows the usage tokens returned by the API, with separate token meters for input, video output, and other output. Check the Pricing table above for current rates.
No. EvoLink provides access via one API key. No Google Cloud project, no Vertex billing, no separate region approval. Same authentication as Veo 3.1 and Seedance 2.0 on EvoLink.
Four modes are available: gemini-omni-flash-text-to-video, gemini-omni-flash-image-to-video, gemini-omni-flash-reference-to-video, and gemini-omni-flash-video-edit. All share the same async video API endpoint.
Yes. Pass a callback_url (HTTPS) when submitting the task and EvoLink can POST task updates to your endpoint when the task reaches a terminal state. Polling the task status endpoint also works if you do not provide a callback URL.
Failed tasks return a failed status with an error reason. For application-level retry, inspect the error, keep the original parameters for debugging, and resubmit only when the input or transient failure mode is clear.
Yes — this is one of Gemini Omni's main workflow differences. Use a natural-language edit instruction and validate how well the selected route preserves the surrounding scene, subject identity, and motion across iterations.
Generation modes support configurable 3-10 second or Auto clips. Auto is estimated as 10 seconds for reservation. Video edit accepts one MP4 input up to 10 seconds. For longer narratives, chain multiple clips using long-context character consistency.
Yes. Pass a reference image URL and Gemini Omni uses it as an identity anchor for the generated video.
Seedance 2.0 has strong benchmark and multimodal reference signals, while Veo 3.1 remains a strong cinematic generation baseline with advanced Flow and extension workflows. Gemini Omni is different because developers are evaluating it for conversational editing, multi-input generation, and short-form iteration.
Yes. EvoLink exposes Gemini Omni, Veo 3.1, Nano Banana 2, and the rest of the Gemini family through a single API key. Switch by changing the model parameter.
Figures are pre-bill estimates. Actual charges follow the upstream usage tokens returned by the model.