クイックアンサー
Grok Imagine is xAI's video generation model, sold on the xAI API as two model IDs: grok-imagine-video at $0.05 per output second and the current grok-imagine-video-1.5 at $0.08. Both bill a flat per-second list rate — the pricing table lists no resolution tiers — so a 6-second 1.5 clip costs $0.48 and a 15-second one $1.20.
Clips run 1–15 seconds at a user-set length in 480p or 720p across seven aspect ratios (16:9 default). One API covers text-to-video, image-to-video where the image becomes the first frame, prompt-based video editing that keeps the original duration, reference-to-video, and extension that continues from the last frame. Requests are asynchronous: submit, poll a request ID, fetch the finished URL — the xAI SDK polls for you.
Video 1.5 (June 2026) is xAI's current version: launch reporting credits it with 24 fps clips, synchronized audio and leaderboard-topping preference results, with 6-second 720p renders at roughly 25 seconds. The consumer surface is the Grok app at grok.com/imagine inside Grok plans; the real per-second price path is the API.
Use the xAI API
Get a console.x.ai key and call grok-imagine-video-1.5 at $0.08 per output second (or 1.0 at $0.05); requests are asynchronous and the SDK polls for you.
Grok Imagine pricingUse the Grok app
grok.com/imagine generates clips inside Grok consumer plans; plan quotas and terms are not verified on this page.
Open Grok ImagineNo local option
Grok Imagine weights are closed and there is no open-weights release; every mode runs through xAI's hosted API.
See access01生成モード
Grok Imagine is hosted-only — every mode runs through the asynchronous xAI API or the Grok app, billed per output second at the model's flat rate.
Text to Video
T2VGenerate from a prompt with configurable duration (1–15 s), aspect ratio and resolution. Requests are deferred: submit, then poll the request ID until done.
text → video
Image to Video
I2VAnimate a still image with a prompt — the source image becomes the first frame of the generated clip.
image + text → video
Video Edit
V2VSend a video with an edit prompt; the model re-renders the clip and keeps its original duration.
video + text → video
Reference to Video
R2VReference images steer the content without forcing the first frame.
images + text → video
Video Extension
EXTContinue a clip from its last frame; the extension is merged into one video.
video → video
02Official pricing & registration
Per-second list prices from xAI's API catalog, checked 2026-09-30. Both video model IDs bill a flat rate per second of generated video with no resolution tier; generation is asynchronous and xAI's SDKs handle polling. API access requires an xAI console account.
| Model ID | Price per output second | What it covers | Source |
|---|---|---|---|
| grok-imagine-video | $0.050 | Video generation: text or image input, 480p/720p, 1–15 s, edit and extension | 公式情報源 ↗ |
| grok-imagine-video-1.5 | $0.080 | Current recommended video model (June 2026): 24 fps, better physics, faster renders | 公式情報源 ↗ |
| Billing basis | Flat per output second | The pricing table lists no resolution tiers; duration is user-set between 1 and 15 seconds | 公式情報源 ↗ |
| Free tier | None documented | xAI's pricing page documents no free video allowance; API use is billed per second from the start | 公式情報源 ↗ |
03プロンプトの書き方
例: A paper lantern drifts over a night market, the camera cranes up slowly, neon reflections ripple across wet asphalt, distant chatter and sizzling woks, warm cinematic realismOpen xAI video docs
04FAQ
Does Grok have a video generation model?
Yes — Grok Imagine is xAI's video model. The current version is Grok Imagine Video 1.5, announced in June 2026, served in the Grok app at grok.com/imagine and on the xAI API as grok-imagine-video-1.5.
How much does Grok Imagine cost?
Flat per output second on the xAI API: $0.05 for grok-imagine-video and $0.08 for grok-imagine-video-1.5, with no resolution tier. A 6-second clip costs $0.30 on 1.0 or $0.48 on 1.5; a 15-second 1.5 clip costs $1.20.
What is the difference between grok-imagine-video and grok-imagine-video-1.5?
Video 1.5 is the June 2026 update: launch reporting credits it with better physics, faster renders (6-second 720p clips in roughly 25 seconds versus over 40) and synchronized audio at 24 fps. The API prices the newer model at $0.08 per second against $0.05 for 1.0.
How long can a Grok Imagine video be?
You set the duration between 1 and 15 seconds at 480p or 720p. Editing a clip keeps its original duration, and extension continues from the last frame with the new segment merged into one video.
Does Grok Imagine generate audio?
Clips carry synchronized sound — music, effects and dialogue per xAI's 1.5 launch reporting. The API documents no audio parameter: audio is built into the output rather than offered as a toggle.
Can I use Grok Imagine for free?
The xAI API documents no free video allowance — every generated second is billed at the list rate. The Grok app includes Imagine inside Grok's consumer plans, but this page does not verify app plan terms; check grok.com for current credits.
05出典
xAI API model catalog
対象範囲： Official model IDs and per-second prices for grok-imagine-video ($0.05) and grok-imagine-video-1.5 ($0.08), checked 2026-09-30.
xAI video generation and edit guide
対象範囲： Official modes (text/image input, edit, reference, extension), 1–15 s durations, 480p/720p resolutions, seven aspect ratios and the async request flow.
確認日 2026-09-30
xAI API pricing
対象範囲： Official per-second pricing table: no resolution tiers and no free video allowance listed (checked 2026-09-30).
Grok Imagine Video 1.5 launch coverage
対象範囲： Third-party reporting of xAI's June 2026 1.5 announcement: 24 fps clips with synchronized audio, faster renders and leaderboard-topping preference results.
確認日 2026-09-30