Use Grok chat models for reasoning, coding and agents — priced about 60% below xAI list price — and Grok Imagine for video and image generation, all through one Kie AI API key with pay-as-you-go billing.
Long-form reasoning, coding and agent workflows with selectable reasoning effort.
Text-to-video, image-to-video and image generation with Grok Imagine.
This table updates automatically as new Grok models launch and are priced on Kie AI.
Explore all AI models on Kie AI →
| Model | Type | Launched | Price | Official price | Status |
|---|---|---|---|---|---|
| Grok 4.7 | Chat | Sept. 2026 | $0.80 / $2.40 per 1M tokens | $2.00 / $6.00 | Latest |
| Grok 4.6 | Chat | Aug. 2026 | $0.80 / $2.40 per 1M tokens | $2.00 / $6.00 | Available |
| Grok 4.5 | Chat | Juli 2026 | $0.80 / $2.40 per 1M tokens | $2.00 / $6.00 | Available |
| Grok Imagine Image 2.0 | Text to Image · Image to Image · Image Editing | Aug. 2026 | from $0.02 per image | $0.06 | Latest |
| Grok Imagine Video 1.5 | Image to Video | Juni 2026 | from $0.012 per second | $0.08 | Available |
| Grok Imagine | Text to Video · Image to Video · Text to Image · Image to Image | Nov. 2025 | from $0.012 per second | $0.05 | Available |
Beyond Grok Imagine, Kie AI serves Grok chat models through the same API key. Here is what the Grok chat API supports today, based on the Kie AI API documentation.
Send chat requests to POST https://api.kie.ai/grok/v1/responses. The endpoint follows the OpenAI Responses format — input and output items plus response.* streaming events — so code written for that format can usually be pointed at Kie AI by changing the base URL, API key and Grok model ID, such as grok-4-7; check the Kie AI API docs for the supported request fields.
Set reasoning effort to low, medium, high or xhigh to trade speed against depth, or leave it unset and let the model choose. Reasoning tokens are billed as output tokens, so a lower effort level typically reduces the cost of each request.
Pass your own tools as function definitions so Grok can call them in agent workflows, and request JSON-schema output in strict mode when your app needs responses it can parse reliably.
Include images by URL or as base64 data in the same request to ask about screenshots, documents or photos. Responses stream by default; turn streaming off to receive the full answer at once after it completes.
For video ideas that start from a written scene, Grok Imagine Text to Video API helps convert prompts into motion-based content. Users can describe the subject, action, setting, camera movement, lighting, mood, and audio direction, while Grok Imagine Video API supports cinematic concepts, product teasers, social media clips, storytelling visuals, and creative testing.
A strong still image can become dynamic video content through Grok Imagine Image to Video API. Developers can animate portraits, product photos, character art, illustrations, concept visuals, and branded assets with motion, depth, camera movement, and scene atmosphere while keeping the original subject, composition, and visual identity as the foundation.
When the goal is still-image creation, Grok Imagine Text to Image API can generate visuals directly from written descriptions. Developers can use Grok Imagine Image API for product concepts, campaign visuals, social media graphics, posters, character designs, creative mockups, and other prompt-based image generation use cases.
For existing visuals that need refinement or transformation, Grok Imagine Image to Image API uses prompt instructions to restyle, adjust, or adapt source images. Developers can use Grok Imagine Image API for photo restyling, image variation, concept refinement, atmosphere changes, creative direction testing, and turning existing assets into new visual styles.
From still images to motion content, Grok Imagine API supports text-to-video, image-to-video, text-to-image, and image-to-image creation. Users can generate original visuals from prompts, transform existing images with instructions, create videos from written descriptions, or animate still images into dynamic clips. This gives users a broader creative range than a single-purpose image or video generator.
For video outputs that need more than silent motion, Grok Imagine Video API can create clips with synchronized audio. Prompts can include ambient sound, music direction, sound effects, or short dialogue cues, allowing the final result to combine motion and sound in one generation process. This is especially useful for cinematic scenes, product showcases, social clips, character moments, and story-driven visual content.
With detailed natural-language instructions, Grok AI API can follow creative direction across image and video generation. Users can describe subjects, actions, visual style, lighting, mood, composition, camera movement, scene atmosphere, and overall creative direction. Better prompt understanding helps different Grok API creation modes produce outputs that feel more controlled and intentional.
Beyond simple image output, Grok Imagine Image API supports both prompt-based image generation and image-to-image transformation. It can create product concepts, posters, social media graphics, character designs, visual mockups, style variations, and refined image concepts from either text prompts or existing visuals. This makes it useful for fast visual exploration, creative refinement, and adapting one visual direction into another.
For different video styles and creative tones, Grok Imagine Video API supports modes such as Normal, Fun, and Spicy. Normal mode is suitable for balanced results, Fun mode can create a more playful interpretation, and Spicy mode may produce bolder stylistic variations. Available modes, supported parameters, and model behavior may change over time, so the latest Kie.ai API documentation should be used as the final reference.
A static product image often needs movement before it feels ready for a landing page, launch teaser, or digital ad. Grok Imagine Video API can add camera motion, lighting shifts, background atmosphere, and product-focused movement to help turn product concepts or existing product photos into short showcase clips. It works well for ecommerce previews, fashion items, packaging visuals, app feature displays, consumer electronics, beauty products, and premium brand assets.
Marketing ideas usually begin as a rough visual direction rather than a finished video plan. With Grok Imagine Text to Video API, users can describe a sale theme, lifestyle setting, brand mood, product placement, seasonal atmosphere, or sound direction, then generate short promotional clips for campaign testing. This is useful for ad concepts, landing-page hero videos, social launch materials, event announcements, and creative variations before a final production direction is chosen.
Character-based content needs motion that keeps the original identity recognizable. Grok Imagine Image to Video API can animate portraits, avatars, mascots, anime-style characters, game concepts, or illustrated figures while preserving the source image as the visual foundation. Users can guide facial expression, body movement, camera behavior, lighting, and atmosphere, making it useful for virtual influencers, digital mascots, profile videos, story previews, and character-driven social content.
Some video ideas are less about selling a product and more about testing a scene. Grok Video API can help explore film-style shots, music video concepts, game cutscene ideas, atmospheric environments, and short narrative moments. A prompt can define the location, action, camera movement, lighting, weather, emotional tone, and audio direction, allowing users to quickly test how a scene might feel before investing in a larger creative production.
When a campaign, event, or launch needs a strong visual starting point, Grok Imagine Image API can help create poster-style drafts and key visual concepts from simple prompts. It works well for promotional posters, music visuals, launch graphics, film-style artwork, seasonal campaign images, and other bold compositions where mood, layout direction, and visual impact matter from the beginning.
For content that depends on a strong first impression, Grok Imagine Text to Image API can generate cover-style visuals from written descriptions. Blog covers, playlist artwork, podcast covers, video thumbnails, social banners, and channel graphics can all begin as prompt-based image concepts, giving users more freedom to test subject placement, color mood, visual focus, and overall creative direction before moving into final design.
In many projects, the first need is not a finished asset but a clearer visual direction. Grok Image API can generate reference images for creative planning, mood exploration, brand direction, character styling, environment concepts, and aesthetic testing. This makes it useful for early-stage ideation, especially when teams want several visual options to compare before choosing a more specific design path.
Rather than rebuilding an image concept from zero, Grok Imagine Image to Image API can transform an existing visual into a new direction while keeping the source image as a reference. Lighting changes, style shifts, background adjustments, color palette updates, and broader visual refinements can all be explored through prompt-based editing, making it useful for asset improvement, image variation, and adapting one visual idea into multiple creative versions.
A strong Grok API prompt should make the main subject easy to understand from the beginning. For images, the subject may be a product, character, poster concept, environment, or visual asset. For videos, it can be a person, object, scene, or action that drives the motion. Clear subject definition helps the output stay focused instead of becoming visually scattered.
Visual direction matters across both image and video generation. Users can improve Grok Imagine API results by describing style, mood, lighting, color tone, composition, and atmosphere in natural language. Instead of writing only “a futuristic city,” a stronger prompt can mention neon lighting, rainy streets, cinematic shadows, reflective surfaces, or a calm cyberpunk mood.
Video generation needs more than a static scene description. Grok Imagine Video API prompts should explain what moves, how fast it moves, and how the camera behaves. Details such as slow push-in, handheld tracking, gentle orbit, dramatic zoom, drifting smoke, flowing fabric, or a character turning toward the camera can help the final video feel more directed.
When using Grok Imagine Image to Video API, the source image already provides the visual foundation. The prompt should focus on motion, camera direction, atmosphere, and sound cues rather than re-describing every object in the image. For better consistency, the requested action should match the subject and composition already visible in the uploaded image.
For image-to-image generation, vague editing requests can lead to unpredictable results. Grok Imagine Image API works better when users describe the exact visual change they want, such as warmer lighting, a luxury product background, anime-style rendering, softer shadows, a cleaner poster composition, or a different color mood. Clear transformation goals help preserve the useful parts of the source image while changing the intended elements.
Small wording changes can create noticeable differences in Grok API image and video results. Users can improve output quality by adjusting one element at a time, such as lighting, camera movement, style, action, or audio direction. This makes it easier to compare results and understand which prompt details have the strongest effect on the final image or video.
Kie.ai provides a practical pricing structure for users who want to test Grok API capabilities before larger-scale usage. You can evaluate text-to-video, image-to-video, text-to-image, and image-to-image results based on available credits, then decide how the API fits your actual content or product needs. This makes cost control easier than committing to a workflow before seeing real output quality.
A reliable API experience depends on more than model quality. Kie.ai offers Grok Imagine API documentation that helps users understand request methods, available endpoints, input requirements, response handling, and usage notes. For details such as supported parameters, mode options, and current model behavior, the latest Kie.ai API documentation should always be used as the final reference.
Kie.ai brings Grok Imagine API into a broader AI generation platform, making it easier to manage image and video creation from one place. Users can work with Grok Image API and Grok Video API capabilities without switching between disconnected tools, which is useful for teams that need both still-image output and motion content in the same creative pipeline.
Before connecting Grok API to a product or workflow, users can test outputs with available credits on Kie.ai. This helps evaluate prompt behavior, generation quality, video motion, image transformation, audio performance, and overall workflow fit. Credit availability and testing scope may vary by account status and current platform policy, so users should check the latest Kie.ai information before planning usage.
Kie.ai supports different Grok Imagine API creation paths, including prompt-based image generation, image transformation, text-to-video generation, and image-to-video animation. For video generation, available modes may include Normal, Fun, and Spicy depending on current API support. This gives users more room to explore different creative directions while keeping the workflow organized under one platform.
AI generation models change quickly, and API behavior may evolve as new versions, modes, or parameters become available. Kie.ai helps users access updated Grok API capabilities, review the latest documentation, and get support when integration or generation issues appear. This is useful for users who need a more stable path for testing, building, and maintaining image or video generation features.
