Mixer AI
Grok 4.7
Explore ideas, analyze images and turn observations into usable writing.
Updated
When to choose it
Use Grok 4.7 in Canvas for a draft, a reference review or checking a text against a brief. Instead of asking what it thinks generally, give criteria: what is visible, what a script is missing or which claims lack support. The answer then has a clear role in the next step of your work.
- Text tasks and image analysis in one request.
- Request editing, a short plan, comparison or a finished prompt.
Modes and settings
- Select Grok 4.7 as a text model in Canvas.
- Up to 5 text blocks and 10 images, producing text without automatic media generation.
Price in coins
| Version | Input, coins / 1M tokens | Cached input, coins / 1M tokens | Output, coins / 1M tokens |
|---|---|---|---|
| Grok 4.7 | 109.2 | 27.3 | 327.6 |
Rates are in coins per million tokens, not per reply. A token is a piece of text, not necessarily a word. Input, cached input and output have separate rates. Canvas reserves an estimate and settles from actual usage; images also affect usage.
≈ marks a fractional rate rounded to two decimal places for display; billing uses full precision.
In US dollars: from ≈ $1.20 per 1M input tokens. 1 coin ≈ $0.011.
What you can make
- Break a visual reference into subject, composition, light and camera treatment.
- Turn several notes into a clear action plan.
How to direct the result
- Separate observations from assumptions; ask it not to present an image-based guess as fact.
- Specify audience, purpose and length. For a prompt, request the final wording without an introduction.
Example: analyze a visual reference
Analyze the connected image to prepare a prompt. Describe only visible subject, composition, light, palette and depth of field. Do not guess the camera brand or exact location. Then write one complete prompt with the same visual mood for a ceramic-cup scene. Stay under 2000 characters.
What to keep in mind
- Using Grok does not automatically provide live X feeds or web search.
- Canvas saves up to 3400 answer characters. Verify facts, numbers and small image lettering.
Sources
Model facts verified: 2026-09-23.
Guides and comparisons using this model
FAQ
Is this a video-generation model?
No. Grok 4.7 on this page returns text. Separate Grok Imagine models create photos and video.
Can I get a prompt from a photo?
Yes. Connect the image and specify what to preserve and what to change in the new description.
Does it automatically know the latest news?
Do not rely on that. Add current verified material directly to the task.
How do I reduce invented details?
Ask it to distinguish observations from assumptions and flag missing information. Check important claims afterwards.
Can I compare several pictures?
Yes, up to ten. Number them and apply common comparison criteria.
Why does the page address contain an older version number?
The address is preserved so existing links continue working. The current model name and settings appear on the page and in Canvas.
What changes the price?
It depends on the version, the text and images supplied, and the response length. Canvas reserves an estimate first, then bills actual usage. The rates in the table help compare versions; they are not a fixed price for every reply.
Do I need a separate model subscription?
No. In Mixer AI, you top up one coin balance and use it for the tasks you need. You do not need a separate subscription to this model.