all tools / vision
As of the latest check, Glimind tracks 72 vision AI agent (MCP) tools, of which 1 are healthy right now. The table below ranks them live-first by real-time reliability.
| tool | status | what it does | |
|---|---|---|---|
| vision.outfit.detect_pieces io.github.rofoso-com/mcp | healthy | Deconstruct an outfit image or influencer photo into individual fashion pieces (e.g. Upper-body garment, Lower-body garment, Dress, Footwear | alternatives |
| vision_query io.github.saloprj/dialogbrain | auth | Look at the screen currently being shared in a meeting and answer a question about it. Returns a natural-language answer based on the visual | alternatives |
| vision io.github.moralito311-andr/andreax | unknown | Computer vision: describe or analyze an image with a LOCAL multimodal model (llava). For agents that need to 'see' (describe scenes, read di | alternatives |
| vision-multi-language io.github.moralito311-andr/andreax | unknown | vision specialized for multi language [x402: 0.01 USDC on Base, pay-per-use] | alternatives |
| vision.block_mapper rogermt/forgesyte | unknown | Block mapping plugin for detecting and mapping structural blocks in images | alternatives |
| vision_describe systemmin/image-mcp | unknown | 生成单张图片的详细文字描述/识别结果 | alternatives |
| vision_analyze systemmin/image-mcp | unknown | 对多张图片综合分析/对比 | alternatives |
| vision_analyze tmchao7/local-vision-mcp | unknown | Analyzes a local image (optionally comparing with a second image) using a local Ollama vision model and returns a structured text report. | alternatives |
| vision_compare groxaxo/DeepSeek Vision MCP | unknown | Compare two or more images | alternatives |
| vision_ocr groxaxo/DeepSeek Vision MCP | unknown | Screenshot/document OCR | alternatives |
| vision_describe groxaxo/DeepSeek Vision MCP | unknown | Faithful scene/UI description | alternatives |
| vision_locate groxaxo/DeepSeek Vision MCP | unknown | Best-effort UI/object localization to normalized coordinates | alternatives |
| vision.ocr rogermt/forgesyte | unknown | Optical character recognition plugin for extracting text from images | alternatives |
| run_vision_agent n8n-nodes-browser-use-orchestrator | unknown | Capture full-page + description par un LLM vision. | alternatives |
| vision_analyze look-at-deepseek | unknown | 图片理解。task 参数内置 6 类任务模板（describe 描述 / chart 图表解读 / ui 界面审查 / formula 公式提取 / translate 翻译 / meme 梗图解读），outputFormat=json 让模型返回对应 JSON 对象（图表→数据 | alternatives |
| vision_image_info look-at-deepseek | unknown | 图片元信息（宽高/格式/大小），本地用 sharp 或魔数头解析、远程仅 HEAD 探测，零 API 成本 | alternatives |
| arachne_visual_snapshot arachne-mcp | unknown | Captures URL → creates baseline (1st time) or compares | alternatives |
| arachne_visual_diff arachne-mcp | unknown | Compares with baseline → diff + semantic verdict | alternatives |
| arachne_screenshot_vision arachne-mcp | unknown | Screenshot + full vision pipeline | alternatives |
| arachne_pdf_vision arachne-mcp | unknown | Extracts images from PDF + vision analysis | alternatives |
| arachne_visual_gates arachne-mcp | unknown | Deterministic audit: overflow, text collision, JS errors | alternatives |
| arachne_visual_report arachne-mcp | unknown | Test suite → HTML side-by-side | alternatives |
| vision_chat maxliven/cn-llm-bridge | unknown | 多轮视觉对话 | alternatives |
| vision_analyze maxliven/cn-llm-bridge | unknown | 图片分析，返回结构化 JSON | alternatives |
| desktop_vision mrpulor-gh/nuphus-mcp | unknown | Send a screenshot to your own vision model via an OpenAI-compatible API for semantic understanding. | alternatives |
| point_objects mseep-moondream-mcp | unknown | Get coordinates of objects in images. | alternatives |
| analyze_image mseep-moondream-mcp | unknown | Multi-purpose image analysis tool. | alternatives |
| describe_image helalove0812-source/VisionPower | unknown | Analyze images via local path, URL, base64, or ordered multi-image array, returning text description. | alternatives |
| vision mingrath/chatgpt-sol-local-bridge | unknown | Takes screenshot and OCR | alternatives |
| ai_vision ocbenji/bitcoinbenji-mcp | unknown | Vision analysis | alternatives |
| ai_ocr ocbenji/bitcoinbenji-mcp | unknown | OCR from images | alternatives |
| vision_describe Seom-ingit/vision_kit | unknown | 看一张图，返回中文描述，或回答关于图的具体问题（传入 prompt） | alternatives |
| vision_extract Seom-ingit/vision_kit | unknown | 结构化识别图中的数字标注：向量 / 矩阵 / 坐标 / 角度，带维度校验与补漏重试 | alternatives |
| vision_qa systemmin/image-mcp | unknown | 针对图片提问，模型按图片内容回答 (VQA) | alternatives |
| see_image PichurChill/j-can-see | unknown | Describe or compare images, zoom into regions | alternatives |
| vision mingrath/chatgpt-mac-bridge | unknown | Take a screenshot and analyze | alternatives |
| look ZYYDI1959/winagent | unknown | 截屏并通过VLM识别屏幕元素位置或回答屏幕相关问题 | alternatives |
| vision_see_screen huocai250/mcp-hands | unknown | See the current screen. | alternatives |
| vision_read_screen_text huocai250/mcp-hands | unknown | Read text from the screen. | alternatives |
| vision_describe_ui huocai250/mcp-hands | unknown | Describe the UI and suggest next action. | alternatives |
| vision_describe blind-vision-mcp | unknown | Analyze an image and return a text description. | alternatives |
| vision_qa karond-is-me/vision-mcp | unknown | 对图片提问，返回基于图片内容的文字回答。 | alternatives |
| vision_describe karond-is-me/vision-mcp | unknown | 详细、真实地解读图片内容，不推测或脑补。 | alternatives |
| vision_ocr karond-is-me/vision-mcp | unknown | 提取图片中所有可见文字。 | alternatives |
| vlm_snapshot desktop-test-mcp | unknown | VLM 从截图识别全部 UI 元素及坐标 | alternatives |
| vlm_click desktop-test-mcp | unknown | 用自然语言描述目标，VLM 定位后在坐标处点击 | alternatives |
| vlm_assert desktop-test-mcp | unknown | VLM 视觉断言，返回 PASS/FAIL | alternatives |
| vlm_navigate desktop-test-mcp | unknown | OODA 循环自动探索，完成多步场景 | alternatives |
| desktop_vision_ask desktop-test-mcp | unknown | 向 VLM 提问关于当前屏幕的问题 | alternatives |
| vision-arwaky rakaarwaky/mcp-arwaky | unknown | Computer vision MCP providing VLM, OCR, and visual memory capabilities. | alternatives |
| vision.diff Aelindra/mobile-agent-harness | unknown | Compares two screenshots | alternatives |
| ui2.state Aelindra/mobile-agent-harness | unknown | Fuses the accessibility tree with OCR for canvas-drawn UIs | alternatives |
| ui2.screen_evidence Aelindra/mobile-agent-harness | unknown | Collects measurable screen features without interpreting them | alternatives |
| ui2.find_templates Aelindra/mobile-agent-harness | unknown | Finds templates in the UI | alternatives |
| vision.ocr Aelindra/mobile-agent-harness | unknown | Performs OCR on the screen | alternatives |
| anchor_visual_environment Alpha-Park/genpark-ambient-vision-anchor-skill | unknown | Indexes ambient visual frames into spatial-temporal scene embeddings with entity object bounding. | alternatives |
| ground_spatial_query Alpha-Park/genpark-ambient-vision-anchor-skill | unknown | Resolves deictic queries ('what is that?', 'summarize this page') against anchored visual coordinates. | alternatives |
| vision bladebro | unknown | Take a screenshot or perform visual analysis. | alternatives |
| vision_ocr smjda0719/dsh-vision | unknown | 文字提取：从图片中提取文字内容。 | alternatives |
| vision.screenshot Aelindra/mobile-agent-harness | unknown | Takes a screenshot | alternatives |
| vision_analyze smjda0719/dsh-vision | unknown | 图片/图表问答：对照片、截图、UI、图表、曲线图、原理图等进行视觉理解与问答。 | alternatives |
| vision_video smjda0719/dsh-vision | unknown | 视频理解：对视频进行问答或总结。 | alternatives |
| mcp_vision_augment_vision CaoMeiYouRen/vision-augment | unknown | 入口：看图问答/OCR/文档解析 | alternatives |
| vision_locate mcp-stealth-chrome | unknown | Locate elements on the page using vision-LLM. | alternatives |
| vision_validate Tyrannmisu/VisionatorMCP | unknown | Check if an object/element is present in the image, optionally cropping to a grid region | alternatives |
| vision_compare Tyrannmisu/VisionatorMCP | unknown | Compare two screenshots with pixel-level diff and VLM semantic analysis | alternatives |
| vision_color Tyrannmisu/VisionatorMCP | unknown | Dominant color extraction or percentage of pixels in custom color ranges | alternatives |
| read_screen orshih6/iphone-mirror-mcp | unknown | Get OCR text with bounding boxes as JSON | alternatives |
| find_text orshih6/iphone-mirror-mcp | unknown | Find text matches with center points | alternatives |
| vision_analyze groxaxo/DeepSeek Vision MCP | unknown | General one/multi-image analysis with a custom prompt | alternatives |
| vision io.github.djautomd-lab/aipaygent | down | Analyze images and provide descriptions or insights. | alternatives |
| ai_vision io.github.Mgnlockii/reelfy | down | Analyze an image with Claude: OCR, chart reading, screenshot explanation. ~$0.05. | alternatives |
As of the latest check, Glimind tracks 72 vision AI agent (MCP) tools, of which 1 are healthy right now. The table below ranks them live-first by real-time reliability. Glimind rates each neutrally (0–100) from safe liveness probes plus privacy-clean real-usage outcomes — see each tool's live page for its current score and a working alternative if it's down.
Ranked live-first by current verdict (healthy → degraded → down), then by the neutral reliability score. Glimind sells no tools, so the ranking is unconflicted.
Every tool here links to its live alternatives — capability-matched substitutes that are healthy now. Or query https://glimind.com/v1/alternatives/{toolId}.
Live data via MCP/REST. Neutral ratings — Glimind only measures.