正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Moondream 3 can parse complex parking signs in one step. Prompt: "extract sign details" → JSON of each rule + transcription. No OCR stack, no regex: just vision that understands structure. ⚡️Fast, cheap, grounded vision AI.

moondream

10,964 subscribers

21,862 次观看 • 1 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

Moondream doesn’t just OCR, it understands. Full JSON receipts, instant Q&A. Open, tiny, blazingly fast.

Moondream doesn’t just OCR, it understands. Full JSON receipts, instant Q&A. Open, tiny, blazingly fast.

moondream

281,016 次观看 • 1 年前

Prime Intellect has acquired Moondream We’re bringing Moondream’s open vision models and inference optimizations into our stack Train and deploy multimodal models across vision, video, voice, decision models and computer use

Prime Intellect has acquired Moondream We’re bringing Moondream’s open vision models and inference optimizations into our stack Train and deploy multimodal models across vision, video, voice, decision models and computer use

Prime Intellect

149,295 次观看 • 29 天前

GutenOCR: A Grounded Vision-Language Front-End for Documents Fine-tunes Qwen2.5-VL-3B and 7B using a grounded OCR evaluation protocol, substantially improving region- and line-level OCR as well as text-detection recall

GutenOCR: A Grounded Vision-Language Front-End for Documents Fine-tunes Qwen2.5-VL-3B and 7B using a grounded OCR evaluation protocol, substantially improving region- and line-level OCR as well as text-detection recall

Niels Rogge

45,811 次观看 • 9 个月前

last week we open sourced: anydoc: parse 14 document formats, ~5ms per page pdf-inspector: parses + classifies pdfs, no waiting on OCR 14k stars each, built in rust. now powering Firecrawl /parse, which adds OCR for the hard ones what should we open source next?

last week we open sourced: anydoc: parse 14 document formats, ~5ms per page pdf-inspector: parses + classifies pdfs, no waiting on OCR 14k stars each, built in rust. now powering Firecrawl /parse, which adds OCR for the hard ones what should we open source next?

Nicolas Camara

101,602 次观看 • 3 个月前

Anthony Evans is just lightning fast. Great feet and downfield vision.⚡️#GoDawgs

Anthony Evans is just lightning fast. Great feet and downfield vision.⚡️#GoDawgs

Aaron Leicht

35,549 次观看 • 2 年前

🚨 Dola AI is giving FREE AI video credits for Seedance 2.5 !! Seedance 2.5 just changed what AI video models can do. No complex setup. You can literally create cinematic short films from just ONE prompt🤯 #dola #ai #seedance2.5

🚨 Dola AI is giving FREE AI video credits for Seedance 2.5 !! Seedance 2.5 just changed what AI video models can do. No complex setup. You can literally create cinematic short films from just ONE prompt🤯 #dola #ai #seedance2.5

SARAH

174,076 次观看 • 3 个月前

Wow!! You can now scrape ANY website by just writing a prompt. Using 's /extract endpoint, just describe what you want to extract in a prompt. This produces LLM-ready structured output. No more hard coding!

Sensitive content

Wow!! You can now scrape ANY website by just writing a prompt. Using 's /extract endpoint, just describe what you want to extract in a prompt. This produces LLM-ready structured output. No more hard coding!

Avi Chawla

250,843 次观看 • 1 年前

Seedance 2.0 is absolutely GOATED. Created an epic Thanos vs Ultraman showdown in minutes, just a prompt and a vision. No VFX skills, no studio, just pure imagination unleashed.

Seedance 2.0 is absolutely GOATED. Created an epic Thanos vs Ultraman showdown in minutes, just a prompt and a vision. No VFX skills, no studio, just pure imagination unleashed.

Duet | AI

18,924 次观看 • 8 个月前

Example of JSON prompt structure to make your own video intros 📽️

Example of JSON prompt structure to make your own video intros 📽️

AlexandrIA

126,246 次观看 • 1 年前

wow.. you can make ads in 1 click now just upload a photo of your product, Higgsfield AI can make studio level ads in mins and.. free actors, free studios, no prompt needed step by step tutorial:

wow.. you can make ads in 1 click now just upload a photo of your product, Higgsfield AI can make studio level ads in mins and.. free actors, free studios, no prompt needed step by step tutorial:

el.cine

79,188 次观看 • 1 年前

Huge leap in prompt following: Imagine 1.0 understands your vision - even with follow up prompts - and brings it to life.

Huge leap in prompt following: Imagine 1.0 understands your vision - even with follow up prompts - and brings it to life.

xAI

254,302 次观看 • 9 个月前

Check this!! Now you can scrape ANY website by just writing a prompt. Using 's /extract endpoint, just describe what you want to extract in a prompt. This produces LLM-ready structured output. No more hard coding!

Check this!! Now you can scrape ANY website by just writing a prompt. Using 's /extract endpoint, just describe what you want to extract in a prompt. This produces LLM-ready structured output. No more hard coding!

Akshay 🚀

64,158 次观看 • 1 年前

It’s AI but no less an amazing vision of Joe Kovacs

It’s AI but no less an amazing vision of Joe Kovacs

Josh (Flyguy)

25,095 次观看 • 3 个月前

A peanut-sized Chinese model just dethroned Gemini at reading documents. GLM-OCR is a 0.9B parameter vision-language model. It scores 94.62 on OmniDocBench V1.5, ranking #1 overall. For context, it outperforms models 100x its size. 100% open-source. It works in two stages. 1. A layout engine detects every region in a document. 2. Each region gets read in parallel. The model predicts multiple tokens per step instead of one. That's what makes it so fast at small size. It handles things most OCR tools struggle with: > Complex tables and nested layouts > Handwritten text and stamps > Math formulas and code blocks > Mixed image-and-text documents You can run it locally through Ollama. It fits on edge devices with limited compute. Every expensive OCR API just got a free competitor.

A peanut-sized Chinese model just dethroned Gemini at reading documents. GLM-OCR is a 0.9B parameter vision-language model. It scores 94.62 on OmniDocBench V1.5, ranking #1 overall. For context, it outperforms models 100x its size. 100% open-source. It works in two stages. 1. A layout engine detects every region in a document. 2. Each region gets read in parallel. The model predicts multiple tokens per step instead of one. That's what makes it so fast at small size. It handles things most OCR tools struggle with: > Complex tables and nested layouts > Handwritten text and stamps > Math formulas and code blocks > Mixed image-and-text documents You can run it locally through Ollama. It fits on edge devices with limited compute. Every expensive OCR API just got a free competitor.

AlphaSignal

92,425 次观看 • 7 个月前

A peanut-sized Chinese model just dethroned Gemini at reading documents. GLM-OCR is a 0.9B parameter vision-language model. It scores 94.62 on OmniDocBench V1.5, ranking #1 overall. For context, it outperforms models 100x its size. 100% open-source. It works in two stages. 1. A layout engine detects every region in a document. 2. Each region gets read in parallel. The model predicts multiple tokens per step instead of one. That's what makes it so fast at small size. It handles things most OCR tools struggle with: > Complex tables and nested layouts > Handwritten text and stamps > Math formulas and code blocks > Mixed image-and-text documents You can run it locally through Ollama. It fits on edge devices with limited compute. Every expensive OCR API just got a free competitor.

A peanut-sized Chinese model just dethroned Gemini at reading documents. GLM-OCR is a 0.9B parameter vision-language model. It scores 94.62 on OmniDocBench V1.5, ranking #1 overall. For context, it outperforms models 100x its size. 100% open-source. It works in two stages. 1. A layout engine detects every region in a document. 2. Each region gets read in parallel. The model predicts multiple tokens per step instead of one. That's what makes it so fast at small size. It handles things most OCR tools struggle with: > Complex tables and nested layouts > Handwritten text and stamps > Math formulas and code blocks > Mixed image-and-text documents You can run it locally through Ollama. It fits on edge devices with limited compute. Every expensive OCR API just got a free competitor.

Jafar Najafov

13,999 次观看 • 7 个月前

this what AI can do in 4K with one prompt in 3 min.. prompt below

this what AI can do in 4K with one prompt in 3 min.. prompt below

el.cine

25,204 次观看 • 6 个月前

Microsoft's OmniParser V2 just dropped. Now AI can : *understands every button* *identifies all interactive elements* *structures the entire GUI* Pure vision-based parsing. GUI automation just got a lot easier 🎯

Microsoft's OmniParser V2 just dropped. Now AI can : understands every button identifies all interactive elements structures the entire GUI Pure vision-based parsing. GUI automation just got a lot easier 🎯

Gradio

62,029 次观看 • 1 年前

This is the moment filmmaking becomes personal for everyone! Kling 3.0, your AI Director understands scene flow, coverage, and cinematic grammar automatically. No manual cutting, no shot stacking — just one prompt, one complete visual narrative.

This is the moment filmmaking becomes personal for everyone! Kling 3.0, your AI Director understands scene flow, coverage, and cinematic grammar automatically. No manual cutting, no shot stacking — just one prompt, one complete visual narrative.

Rishabh

462,787 次观看 • 9 个月前