Google● Available

Google Gemini Omni 1.1 Flash API

Google Gemini Omni 1.1 Flash 视频生成模型，支持 3～10 秒文生、首尾帧图生、多素材参考、视频编辑和视频延长，输出覆盖 360p、720p、1080p 和 4K。

Text to videoImage to videoReference to videoVideo editingVideo extension 360p720p1080p4k 3–10s / auto

What is Gemini Omni 1.1 Flash

Gemini Omni 1.1 Flash is the 1.1-generation Flash route of the Gemini Omni family (the Roko catalog cites the upstream EvoLink model page as its capability source). It adds a video-extension route, widens output from 720p-only to 360p/720p/1080p/4K, and loosens the reference-asset rules. Roko exposes five routes through the unified async video task API, mapped to gemini-omni-1.1-flash-* submodels.

Billing matches the first generation: token-based at $1.43/M input, $16.50/M video output, and $8.47/M other output. Costs float with each task's usage and cannot be quoted precisely before submit; auto durations are model-decided with reservations estimating 10 seconds, settled on the terminal state.

Durations run 3–10 seconds (default 5) at 16:9, 9:16, or auto, with audio always generated. The edit and extension routes accept only auto aspect ratio and can pick auto or one of the four resolution tiers; extension appends a new segment to existing footage, which is the key building block for multi-shot narratives.

Model pricing

Pricing follows the current production configuration and the billing dimensions supported by each route.

Route	Price	Billing rule
Gemini Omni 1.1 Flash	Calculated from task parameters on the pricing page	输入 token $1.43/M；视频输出 token $16.50/M；其他输出 token $8.47/M

Cost estimate examples

Token billing has no fixed unit price; examples demonstrate the method only. Assumed volumes are not measured — actual charges follow task usage and the pricing page.

Formula

`input tokens/1M × $1.43 + video-output tokens/1M × $16.50 + other-output tokens/1M × $8.47`

Same rates as the first generation. Resolution and duration drive consumption; exact reconciliation lands at the task's terminal state.

Worked assumption (not measured)

`assume 3,000 input + 90,000 video-output + 500 other → 0.0043 + 1.485 + 0.0042 ≈ $1.49`

Assumptions illustrate the method only. Consumption differs significantly across 4K and 360p — calibrate per tier before batch runs.

Model information

Only provider-documented routes, capabilities, and limits implemented by this gateway are listed.

Text to video

`gemini-omni-1.1-flash-text-to-video`

Limits: —
Duration: 3–10s / auto

Image to video

`gemini-omni-1.1-flash-image-to-video`

Limits: Images 1–2
Duration: 3–10s / auto

Reference to video

`gemini-omni-1.1-flash-reference-to-video`

Limits: Images 0–10 · Videos 0–3
Duration: 3–10s / auto

Video editing

`gemini-omni-1.1-flash-video-edit`

Limits: Images 0–10 · Videos 1
Duration: 3–10s / auto

Video extension

`gemini-omni-1.1-flash-video-extend`

Limits: Images 0–10 · Videos 1
Duration: 3–10s / auto

Capability source ↗

Key capabilities

Five routes covering generation and post

text-to-video, image-to-video (1–2 images), reference-to-video (up to 10 images + 3 videos), video-edit (1 clip), and video-extend (1 clip) share one task interface — switching is a model-parameter change.

Video extension (video-extend)

Submit existing footage and the model continues from the final frame, turning 3–10-second shots into multi-segment narratives. Continuity varies by source and must be validated on your own shots.

Resolution tiers from 360p to 4K

Prototype at 360p to control cost, deliver at 1080p/4K; the edit and extension routes also support auto (follow the input). Token consumption differs per tier — usage governs billing.

Videos accepted as references

Unlike the first generation's image-only references, 1.1 accepts up to 10 images plus 3 videos, suiting composites that need motion or camera references.

Main parameters

Parameters come from structured model configuration; unsupported or unconfigured fields are omitted.

modelstring

Submodel ID

Default:

gemini-omni-1.1-flash-text-to-video / gemini-omni-1.1-flash-image-to-video / gemini-omni-1.1-flash-reference-to-video / gemini-omni-1.1-flash-video-edit / gemini-omni-1.1-flash-video-extend

promptstring

Google Gemini Omni 1.1 Flash 视频生成模型，支持 3～10 秒文生、首尾帧图生、多素材参考、视频编辑和视频延长，输出覆盖 360p、720p、1080p 和 4K。

Default: Required

durationinteger

Output video seconds

Default: 3–10s / auto

qualityenum

Current billing

Default: 360p / 720p / 1080p / 4k

aspect_ratioenum

Capability

Default: auto / 16:9 / 9:16

image_urlsarray

Image to video: 1–2 · Reference to video: 0–10 · Video editing: 0–10 · Video extension: 0–10

Default: Varies by route

video_urlsarray

Reference to video: 0–3 · Video editing: 1 · Video extension: 1

Default: Varies by route

Capabilities and usage guidance

Key features

Text to video · gemini-omni-1.1-flash-text-to-video
Image to video · gemini-omni-1.1-flash-image-to-video
Reference to video · gemini-omni-1.1-flash-reference-to-video
Video editing · gemini-omni-1.1-flash-video-edit
Video extension · gemini-omni-1.1-flash-video-extend
360p / 720p / 1080p / 4k output
3–10s / auto Duration
Generated audio output
Audio-capable workflow

Limits

Image to video: Images 1–2
Reference to video: Images 0–10 · Videos 0–3
Video editing: Images 0–10 · Videos 1
Video extension: Images 0–10 · Videos 1
Capability source: https://evolink.ai/zh/gemini-omni-1-1-flash

Limits and caveats

Token billing prevents precise pre-submit quotes; auto durations reserve as 10 seconds and settle at the terminal state.
Edit and extension routes accept only auto aspect ratio; resolution can be auto or one of four tiers.
Audio is always generated and cannot be turned off; pick another model when silence is required.
No dedicated negative-prompt parameter — write exclusions into the prompt body.
Upstream output links expire after 24 hours; bake rehosting into the pipeline.
Family and capability claims come from the upstream EvoLink page (the catalog's cited capability source), not from a Google page.

How to call it

POST /v1/videos/tasks with model set to the route's submodel (gemini-omni-1.1-flash-text-to-video / image-to-video / reference-to-video / video-edit / video-extend), plus prompt, duration (3–10 or auto, default 5), resolution, and aspect_ratio.
Attach media per route: 1–2 images for image-to-video; up to 10 images + 3 videos for reference-to-video; one video each for edit and extension via video_urls.
Poll GET /v1/tasks/{task_id} or receive callback_url until succeeded / failed.
On success, read the video URL and rehost within the 24-hour validity window.
On failure, read the reason and resubmit; write exclusions into the prompt (no separate negative-prompt parameter). A successful submit does not guarantee a successful generation.

Typical use cases

Cheap drafts to 4K deliverables

Input: storyboard prompts; action: proof compositions at 360p, then deliver at 1080p/4K; expectation: the same prompt reused across resolutions; limit: token use rises with resolution — calibrate both tiers.

Extending short shots into longer narratives

Input: the previous segment; action: iteratively extend via video-extend; expectation: style and subject carry across segments; limit: each segment is still 3–10 seconds — long films need many extensions and manual joins.

Composites with motion references

Input: character/scene images plus motion or camera-move videos; action: reference-to-video (≤10 images + ≤3 videos); expectation: static assets composited under dynamic guidance; limit: consistency gets harder with more assets — validate per combination.

Choosing against other video models on Roko

A coarse split by billing model and workflow; final decisions should rely on each catalog and your own tests.

vs first-generation Gemini Omni

1.1 adds video-extend, 360p–4K tiers, video references, and a 5-second default; the first generation suffices for 720p-only shorts with no extension.

vs Veo 3.1

Veo has precomputable per-second pricing, n≤4 batches, and 1–3 image references; 1.1 bills in tokens but accepts video references and extension — more flexible workflows, reconciled from usage.

vs per-request channels (Grok)

Grok's fixed per-request pricing is spec-bound; 1.1 trades cost variance for extension, editing, and multi-asset flexibility.

Frequently asked questions

How do I call Gemini Omni 1.1 Flash?

Submit via POST /v1/videos/tasks with model set to the route's gemini-omni-1.1-flash-* submodel, then poll the task to a terminal state before reading results.

How much does one video cost?

Token billing: $1.43/M input, $16.50/M video output, $8.47/M other output; actual cost depends on task usage and cannot be quoted per clip.

Which resolutions and durations are supported?

360p/720p/1080p/4K (edit and extension also support auto), and 3–10 seconds (default 5) or auto. Subject to the current Roko catalog.

How does it differ from first-generation Gemini Omni?

1.1 adds a video-extend route, 360p–4K tiers, video references (up to 3), and a 5-second default; the first generation is 720p-only with no extension and a 10-second default.

How does video extension work?

Hand existing footage to the video-extend route via video_urls; the model continues from the final frame. Continuity must be validated on your own assets, and the aspect ratio is fixed to auto.

How many reference assets are allowed?

The reference route takes up to 10 images plus 3 videos; image-to-video takes 1–2 images. Subject to the current Roko catalog.

Does 4K cost much more?

4K consumes more tokens than lower tiers and settles higher via usage; the exact delta requires measurement, and this page makes no unverified multiplier claims.

How long do output links stay valid?

Upstream links are valid for 24 hours; rehost files to your own storage as soon as the task completes.

Sources and verification scope

Capability claims follow the upstream EvoLink page cited by the Roko catalog; callable routes, parameters, and prices come from the current production catalog. Reviewed: 2026-10-08.

EvoLink: Gemini Omni 1.1 Flash model page Upstream provider page cited by the Roko catalog as capability source
EvoLink docs: Gemini Omni Flash API manual Upstream API conventions shared by the Omni Flash series
Roko API video task docs Async submit, task query, and billing lifecycle on Roko
Roko API model catalog Current routes, prices, and generation config

Models in this family

Explore other available variants from the same model family.