📰 What's New
Toolbox Workspace — 70+ tools ready to use
Search or open a free tool now. When your agent is ready, use the same surface through REST and MCP.
- 70+ tools — image, video, PDF, audio, data, dev, security, agent utilities
- One API key — Claude, Hermes, curl, cron — same surface
- MCP + REST — install once, call any tool from any agent
- Free to try — no signup for free tools · 10/day, 25 with login
Free starters · no credit card
Also try · products beyond the toolbox
From first result to production call
Try it here. Automate the same tool when it works.
No new product to learn: create a free key, copy one request, and keep the same REST or MCP surface as you grow.
- 1Run a free toolConfirm the input and output in your browser.
- 2Create a free key25 daily calls; no card required.
- 3Make the first callCopy curl, JavaScript, or Python and verify the response.
Press ⌘K — try "resize", "short url", "transcribe", or "api key"
Fix the image, merge the PDF, make the QR — no signup for free tools. Same engine your agents use. No login 10/day · Free login 25/day · Pro $5 1,000/day
One mt_ key — every tool via REST or MCP.
Hermes, Claude, cron, or a one-line curl.
Provision
Agent can invite a human — they set a password, you get the key.
✦ Generate a hero → resize → WebP — one workflow, no Figma export dance
✦ Agent screenshots a URL → extracts the article → pastes a summary
✦ Cron scrapes → Hermes processes → Discord gets the file
✦ Voice note in → Whisper STT → structured notes out
🖼️ Media & Files
Pick a category to get started. Each opens a focused workspace with clear steps.
🔤 Text & Data
Text utilities, structured data, and agent-friendly sharing tools.
🛠️ Dev & Security
Everyday developer utilities — no setup required.
🤖 Agent & API
Connect agents and try the REST API interactively.
🎬 Video Tools
Trim clips, GIF, thumbnails, audio extract, metadata, or AI video generation. Pick a mode below.
✨ AI Video Generation
Generate a short clip with Wan 2.5. Use a template + style chips for better results. Takes 1–3 minutes.
curl -X POST https://agentmediatools.com/api/describe-image -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"url":"https://example.com/photo.jpg","task_type":"caption"}'Use the Agent Media Tools API to describe any public image with vision ai (moondream). Access 'AI Image Description API for Agents — Vision Caption & VQA | MCP + REST' at https://agentmediatools.com/tool/describe-image. Here is the API endpoint: curl -X POST https://agentmediatools.com/api/describe-image -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"url":"https://example.com/photo.jpg","task_type":"caption"}' Integrate this tool into your AI agent workflows for seamless automated processing — no account needed for free-tier tools.What you can build
Turn this tool into a repeatable workflow
- ✓ Caption product photos for catalogs and moderation queues
- ✓ Answer visual questions from an agent without a local vision model
- ✓ Feed image understanding into prompt or workflow steps
Start now
Test it here, then connect your agent
Use the working browser tool above, create a free API key, and keep the same integration as your workflow grows.
Access / cost: Always 2 flexible credits · no free FAL calls · Pro monthly credits apply first · MCP tool id describe_image
Frequently asked questions
How much does image describe cost?
Always 2 flexible credits per call. There is no free FAL allotment because each request uses an upstream vision model. Pro and Builder monthly credits apply first.
Can agents call it via MCP?
Yes. Use the describe_image MCP tool or POST /api/describe-image with a public image_url.
Caption vs query?
task_type caption summarizes the image; query answers a prompt/question about the image.
