🤖 Overview

probe-image-size is a lightweight, open-source HTTP agent built upon the probe-image-size Node.js library (maintained by the nodeca organization on GitHub) and used by numerous web services, content management systems, and social media platforms to efficiently retrieve image dimensions without downloading the full file. The agent is not a standalone product but a modular component embedded in broader crawling or preview pipelines.

🌐 Technical Behavior

The agent primarily issues HTTP HEAD requests or GET requests with a Range header (e.g., bytes=0-) to fetch only the image header bytes, typically the first 2–32 KB. This allows it to determine width, height, and format (JPEG, PNG, GIF, WebP) while consuming minimal bandwidth. Crawl frequency depends entirely on the parent application; some implementations send bursts of 10–50 requests per second, while others throttle to 1–2 requests per second. The agent does not follow links or parse HTML — it targets specific image URLs supplied by the calling service. IP ranges are not fixed; they inherit from the hosting environment (e.g., cloud provider or content delivery network) of the service deploying it. The library supports HTTP/1.1 and HTTPS, and respects Accept-Encoding but does not cache results.

📋 robots.txt Compliance

The library itself does not implement robots.txt parsing; however, the parent application using the probe-image-size agent typically respects robots.txt restrictions before generating image URLs. Official documentation on GitHub (https://github.com/nodeca/probe-image-size) does not mention compliance by default, and the agent is designed for server-side use where the calling service usually handles crawl policy. As a result, compliance varies — many services respect Disallow directives for image paths, but others may ignore them for private API endpoints.

🔍 Detection Indicators

The most reliable identifier is the User-Agent string, commonly formatted as probe-image-size/7.2.3 (version varies). Some implementations append a parent agent identifier, e.g., Mozilla/5.0 (compatible; MyService/1.0; +https://example.com) probe-image-size/7.2.3. Behavioral fingerprints include HEAD-first then GET for JPEG dimensions, and a tendency to send multiple requests for different image dimensions (thumbnail, full) within milliseconds. The Accept header is usually missing or set to image/*.

📊 Data Usage

Collected data — specifically image width, height, and format — is used solely for UI rendering decisions, such as generating proportional previews, cropping, or lazy-loading thresholds. The agent does not store or transmit pixel content or metadata beyond the header. The parent service may log the metadata briefly for analytics (e.g., image load times) but typically discards it after processing. No AI training or search indexing is performed by the agent itself; it feeds data into real-time rendering engines.

⚙️ Rate Limiting Policy

Because the agent can generate many rapid requests when a service crawls image-heavy pages, administrators often rate-limit it to 5–10 requests per second per IP to prevent resource exhaustion on origin image servers. This threshold-based blocking is standard for legitimate but aggressive agents that do not require full image downloads, ensuring fair access for all bots.

ⓘ Data Notice: The information presented above has been compiled from publicly available internet sources. Boteraser aggregates this data solely for informational purposes and does not independently classify, evaluate, or endorse any findings about the bots listed. The accuracy and completeness of this information is the sole responsibility of the original publishers. Boteraser and its operators accept no liability for any decisions made based on this data.