Image resolution

The SUPERWISE® Sentinel image resolution guardrail checks the dimensions of images in a prompt and blocks the request when an image falls outside the resolution limit you set. Unlike Sentinel's other guardrails, it works on images, not text — it inspects the images attached to a multimodal prompt before they reach the model.

Use it to keep out images that are too large (which drive up token cost and latency) or too small to be useful, and to reject anything the gateway can't read as an image.

Action: This guardrail only blocks — there's no content to redact, so there's no remediate option. When an image is outside your limit, the request is rejected with the Error message you set and never reaches the model. See Manage guardrails.

Parameters

• Limit — whether the value is a maximum (block images larger than it) or a minimum (block images smaller than it).
• Width × height — the resolution limit, in pixels. An image is blocked when either side crosses the limit: with a maximum, when its width or height is larger; with a minimum, when its width or height is smaller.

An image the gateway can't decode is also blocked.

Example interception

With a maximum of 1024 × 1024 px and a prompt that attaches a 4000 × 3000 px photo:

Result: the image is above the maximum resolution, so Sentinel blocks the request and returns your error message. The image never reaches the model.

Get Sentinel: Sentinel on superwise.ai.

See also

• Manage guardrails: add the guardrail and test it before publishing.
• Language check: block prompts that aren't in your allowed languages.
• Guardrails: what each guardrail catches.