Get top-k semantically unique images (URLs or base64-encoded) using Jina CLIP v2 embeddings and submodular optimization. Use this when you have many visually similar images and want the most diverse subset.
AI agents call deduplicate_images to retrieve information from Jina AI without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
k | number | — | Number of unique images to return. If not provided, automatically finds optimal k by looking at diminishing return |
images | array | — | Array of image inputs to deduplicate. Each item can be either an HTTP(S) URL or a raw base64-encoded image string (without data URI prefix). |
Parameters from the server's own tool schema.
This tool analyzes and processes images to identify semantic uniqueness and return a diverse subset. It does not create, modify, delete, execute code, or trigger external operations. The operation is deterministic data retrieval/filtering with no ability to alter state or cause irreversible changes.
From the tool's definition Tool performs 'Get top-k semantically unique images' using embeddings and optimization—a retrieval and filtering operation with no side effects. It accepts images (URLs or base64) as input and returns a deduplicated subset, which is a read-only transformation.
Documented attack patterns abuse exactly the kind of access deduplicate_images gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Jina AI, and nothing reaches the server without passing your rules. This is the rule we recommend for deduplicate_images:
{
"version": "1",
"default": "deny",
"tools": {
"deduplicate_images": {}
}
}deduplicate_images is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Get top-k semantically unique images (URLs or base64-encoded) using Jina CLIP v2 embeddings and submodular optimization. Use this when you have many visually similar images and want the most diverse subset. It is categorised as a Read tool in the Jina AI MCP Server, which means it retrieves data without modifying state.
deduplicate_images accepts 2 parameters: k, images. The full parameter table on this page comes from the server's own tool schema.
Register the Jina AI MCP server in PolicyLayer and add a rule for deduplicate_images: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Jina AI. Nothing to install.
deduplicate_images is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the deduplicate_images rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for deduplicate_images. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
deduplicate_images is provided by the Jina AI MCP server (jina). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Jina AI, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
21 Jina AI tools catalogued and risk-classified — across an index of 43,000+ MCP servers.