Best for
- Social and ad creative variations
- Prompt-to-image prototypes
- Reference-image edits when the current route supports them
Image API · xAI · creator workflows
Grok Imagine Image API: Fast Image Generation Access
A practical hub for developers who need fast Grok image generation, prompt variations or reference-image experiments without changing the rest of an OpenAI-style integration.
When this model is a good fit
Grok Imagine Image is useful when the output needs fast creative iteration rather than a long editorial workflow. Keep the prompt, image options and returned asset URL in your acceptance record.
Check before production
- Confirm the exact model ID and image parameters
- Open the returned image before counting the request as successful
- Reconcile image count and balance change
What this page does not guarantee
Do not assume every text-model option works on an image endpoint. Capability, resolution and content rules are model-specific and can change.
How to choose this hub
Choose this hub when image iteration speed and creative variation matter more than a fixed editorial style. Compare it against another image route with the same prompt, output size and review rubric; do not decide from a single attractive sample.
| Option to compare | Choose it when | Measure before switching |
|---|---|---|
| Gemini 3.1 Flash Image | The prompt contains detailed product constraints or multimodal input. | Instruction following, reference-image handling and final image cost |
| A text-only model | You only need copy, planning or metadata rather than pixels. | Total workflow cost after removing the image-generation step |
| A second Grok image route | You need to test a different quality or speed tier. | Asset usability, latency distribution and repeatability |
Production workflow
- Write one fixed prompt and define what makes an image acceptable.
- Run a single image request and save the model ID, options and request ID.
- Open or download the asset before marking the generation successful.
- Repeat only after the ledger entry and the acceptance notes are complete.
Pricing and billing
The public Chinese catalog may show a reference per-image price for Grok image routes. Treat that number as a planning reference, not a guaranteed USD settlement rate.
Use a small batch first. Record model ID, image count, response status, request ID and the final ledger entry. A failed attempt and a second successful attempt are separate events.
Minimal API request
Use the shared gateway and keep the model ID in configuration. Start with a small request, save the request ID, and compare usage with the dashboard before increasing concurrency.
curl https://www.gpt345.com/v1/images/generations -H "Authorization: Bearer $GPT345_API_KEY" -H "Content-Type: application/json" -d '{"model":"grok-imagine-image","prompt":"A clean studio product photo of a glass bottle on a white set","size":"1024x1024"}'Acceptance checklist
- The response contains an image result or a clear provider error.
- The output can be downloaded or opened from the returned URL.
- The dashboard charge matches the number of successful image tasks.
Frequently asked questions
What is the model ID?
Use grok-imagine-image exactly as shown in the current model hub.
Can I use one key for text and image models?
Main-site models can share a key when the account and model group allow it, but keep image and text ledger checks separate.
Is this the Seedance site?
No. Grok image access is on the main site. Seedance has a separate site and separate account system.