Image Metadata & AI Analyzer
Extract image metadata, generate AI-powered captions, tags, and keywords. Supports Ollama, LM Studio, OpenAI, and custom endpoints.
How to use this tool
- Upload an image or paste an image URL. Its size, dimensions and aspect ratio appear straight away.
- Choose an AI provider. Google Gemini is free to try with no key; Ollama and LM Studio keep the image on your computer.
- Click "Analyze with AI" to get a caption, tags and the top supertags.
- Copy the results and edit anything the model could not know.
Click to upload or drag and drop
You can also paste an image from clipboard (Ctrl+V)
Upload an image to get started
Load an image from URL, upload a file, or paste from clipboard. Then use AI to generate captions and keywords.
AI Provider Options
Ollama: Free, runs locally. Install from ollama.com, then run: ollama pull gemma3
Google AI: Uses Gemini Flash-Lite. Requests go through this site's server to Google.
LM Studio: Download LM Studio, load a vision model (e.g., Gemma 3 or Qwen2.5-VL), enable CORS, and start the server
OpenAI: Get an API key from platform.openai.com
How the Image Metadata & AI Analyzer works
Drop in an image (up to 20 MB) or load one from a URL. The tool reads its basic properties in your browser: width and height in pixels, aspect ratio (reduced to a simple ratio such as 16:9 where possible), file size and file type. It does not read or display EXIF camera data such as GPS location or camera model.
For the AI part, the image is sent to a vision model of your choice with one fixed instruction: write a one or two sentence caption, list 10 to 15 relevant tags, and pick the 4 or 5 most important ones as "supertags". The model answers in JSON, which the tool parses and displays so you can copy the caption and tags.
You choose where the image goes. Google AI (Gemini Flash-Lite) is the default and needs no key: the image is sent through this site’s server to Google’s Gemini API. Ollama and LM Studio run a model on your own computer, so the image never leaves your machine. OpenAI and any OpenAI-compatible custom endpoint are called directly from your browser with your own API key.
When to use it
- Writing alt text for images on a blog or website, which helps accessibility and image search.
- Drafting titles, captions and keywords for stock photo submissions.
- Tagging a folder of product or reference photos so they are easier to search later.
- Checking an image’s exact dimensions and aspect ratio before uploading it somewhere with size rules.
Tips
- For private or sensitive photos, use Ollama or LM Studio so the image stays on your computer.
- To use Ollama from this page, start it with OLLAMA_ORIGINS set to allow this site, and pull a vision model such as gemma3. In LM Studio, load a vision-capable model and enable CORS in the local server settings.
- Treat AI captions as a first draft. Check names, places and anything the model could not know from the pixels alone.
- If loading from a URL fails, the other site is probably blocking cross-site requests. Download the image and upload it instead.
Frequently asked questions
Is my image stored?
The image file itself is not stored. With the default Gemini option, the image passes through this site’s server to Google, and the server logs basic request details (image size and type, response time, your IP address and browser, and the generated caption and tags) to monitor usage, apply the daily free limit and fix errors. On Google’s free tier, Google may use submitted images to improve its products, so use Ollama or LM Studio for private photos. With Ollama or LM Studio the image never leaves your device; with OpenAI or a custom endpoint it goes directly from your browser to that provider.
Does it show EXIF data or GPS location?
No. It shows dimensions, aspect ratio, file size and file type only. It does not read EXIF fields.
Which provider gives the best captions?
Hosted models such as Gemini and GPT usually give the most accurate, detailed captions. Local models are private and free to run but vary with the model you load; gemma3 is a good starting point in Ollama.
Why did the analysis fail?
You may have reached the daily limit for free analyses, or the shared free Gemini quota may be used up. In either case you can enter your own free Google AI key. For local providers, check that the app is running and allows requests from this site. Very large images can also time out, so try a smaller copy.
Can I use the tags for stock photography?
Yes, as a starting point. Stock sites reward specific, accurate keywords, so remove anything generic or wrong and add the details only you know, such as the location or event.