New to VisionScaper?
Start here
Help Center › API keys & connections › Vision model setting
Vision model setting
The Vision model box chooses which language and vision model fal.ai runs for identify, helpers and checks.
VisionScaper sends many “look at this picture and tell me” jobs, and many text jobs, through fal.ai’s any-llm service. The Vision model (fal any-llm) setting picks the model name.
Before you start
- A saved fal.ai key (see fal.ai key).
Steps
- Open Settings → VisionScaper in WordPress admin.
- In the AI section, find Vision model (fal any-llm).
- The default is
google/gemini-2.5-flash. - To change it, type a model name that fal.ai’s
any-llmservice accepts. Look up the current list in fal.ai’s own documentation. - Press Save Changes.
- Test one feature that uses it, such as identifying a plant or garden item.
What uses this setting
The same model name is sent for:
– Identifying garden items and materials.
– AI Marketing drafts.
– Community picture checks.
– AI quote and property-walk helpers.
– Other AI helper jobs in the Contractor Hub.
Tips
- Leave the default unless you have a clear need.
- Change one thing at a time and test straight after.
- If you break it, paste the default back in.
- The box must hold a plain model name. Special characters are stripped.
If something goes wrong
- Helpers return errors after you change it: the name may not be a valid fal.ai model name. Put
google/gemini-2.5-flashback. - Do not leave the box empty. Paste
google/gemini-2.5-flashback if you are unsure. - Everything AI fails: check the fal.ai key and your balance first.