🤖 AI Tools | Tech Tips | Productivity 🚀 Helping you work smarter 📚 Daily AI tricks & useful tools HIGGS-KG3D1 DM/mail for collabs📩: [email protected]
Joined November 2025
- Tweets4.2K
- Following62
- Followers9.5K
- Likes2.8K
Neo retweeted
The launch had the numbers. the BTS has the personality. you need both to understand what @tavus built
Neo retweeted
Having sex as a wife in the 90s:
Wife: The kids are asleep. Come to bed, darling.
Husband: Give me two minutes. Let me finish watching the match.
Having sex as a wife in 2026:
Neo retweeted
More time to experiment can help you discover your own creative style. Flova 1.5 offers a second year of Pro free during its launch promotion.
Seedance 2.5 720p starts at $0.025/sec.
flova.ai/?refCode=3BFU8P57
🤝 Paid partnership
Neo retweeted
OCR is only useful when you can trust what it extracts.
MiniCPM-OCR is tackling that challenge with a 1B model for document parsing and a separate model to evaluate the results.
From tables and formulas to complex documents, this open-source suite is built to make document processing faster and more reliable.
Your OCR says it’s done—but did it get everything right? 🤔 In high-stakes workflows, even a small error can be costly.
🚀 Meet the MiniCPM-OCR suite: an open-source solution to parse fast and guard the results.
Two new models:
🔹 MiniCPM-OCR 1.0 — SOTA OCR performance and speed at 1B scale. Compact, accurate, and fast.
🔹 MiniCPM-OCR Jev — A 2B decision model that helps assess parsing accuracy, with performance approaching that of Qwen3.5-397B-A17B on JevOCRBench.
⚡ What’s new?
Small model, SOTA results: MiniCPM-OCR 1.0 scores 93.68 on Real5-OmniDocBench and 54.7 on OCRBench v2.
A sharp goalkeeper 🛡️: MiniCPM-OCR Jev is model-agnostic, checking text, formulas, and tables from different OCR parsers. It achieves 93.6 on JevOCRBench, outperforming JEV-35B-A3B (90.1) and approaching Qwen3.5-397B-A17B (93.8).
Up to 128 pages per minute 🤯: With DSpark, MiniCPM-OCR runs at 1.49× the speed of PaddleOCR-VL-1.6 and 3.62× that of MinerU2.5 Pro. The SDK’s concurrency optimizations further boost throughput to as much as 128 pages per minute.
Deploy where you work: Day-0 support for Arm, Houmo AI, Intel, and NVIDIA, plus 10 more chip vendors through FlagOS.
Ready for your stack: vLLM, SGLang, llama.cpp, and Ollama, with quantized models, fine-tuning support, and Agent Skills.
🔗 GitHub: github.com/OpenBMB/MiniCPM-O…
🤗 Hugging Face:
MiniCPM-OCR 1.0：huggingface.co/openbmb/MiniC…
MiniCPM-OCR Jev：huggingface.co/openbmb/MiniC…
👉 Try the Demo：platform.modelbest.ai/consol…
Use the API：platform.modelbest.ai/consol…
Neo retweeted
🏆 NVIDIA Nemotron 3.5 Transcribe debuts at #1 among 14 systems on Voice Arena’s Diarization + ASR leaderboard.
Across the full benchmark, it achieves 32.43% cpWER.
This track tests whether a system can get both the words and the speakers right. We evaluated 139 held-out conversations with 2-8 speakers across in-person and online settings, including overlapping speech.
Nemotron 3.5 Transcribe is first in both settings: 34.08% cpWER in person (far field) and 22.13% on online calls.
It has the lowest word deletion rate on the leaderboard. Deletions are 8.32% of reference words, against a median near 31%.
Congratulations to @NVIDIAAI on the release ! 👏
Neo retweeted
Your OCR says it’s done—but did it get everything right? 🤔 In high-stakes workflows, even a small error can be costly.
🚀 Meet the MiniCPM-OCR suite: an open-source solution to parse fast and guard the results.
Two new models:
🔹 MiniCPM-OCR 1.0 — SOTA OCR performance and speed at 1B scale. Compact, accurate, and fast.
🔹 MiniCPM-OCR Jev — A 2B decision model that helps assess parsing accuracy, with performance approaching that of Qwen3.5-397B-A17B on JevOCRBench.
⚡ What’s new?
Small model, SOTA results: MiniCPM-OCR 1.0 scores 93.68 on Real5-OmniDocBench and 54.7 on OCRBench v2.
A sharp goalkeeper 🛡️: MiniCPM-OCR Jev is model-agnostic, checking text, formulas, and tables from different OCR parsers. It achieves 93.6 on JevOCRBench, outperforming JEV-35B-A3B (90.1) and approaching Qwen3.5-397B-A17B (93.8).
Up to 128 pages per minute 🤯: With DSpark, MiniCPM-OCR runs at 1.49× the speed of PaddleOCR-VL-1.6 and 3.62× that of MinerU2.5 Pro. The SDK’s concurrency optimizations further boost throughput to as much as 128 pages per minute.
Deploy where you work: Day-0 support for Arm, Houmo AI, Intel, and NVIDIA, plus 10 more chip vendors through FlagOS.
Ready for your stack: vLLM, SGLang, llama.cpp, and Ollama, with quantized models, fine-tuning support, and Agent Skills.
🔗 GitHub: github.com/OpenBMB/MiniCPM-O…
🤗 Hugging Face:
MiniCPM-OCR 1.0：huggingface.co/openbmb/MiniC…
MiniCPM-OCR Jev：huggingface.co/openbmb/MiniC…
👉 Try the Demo：platform.modelbest.ai/consol…
Use the API：platform.modelbest.ai/consol…
Neo retweeted
The ability to manage several conversations within one project is a practical improvement for anyone working through complex creative ideas. Add better asset organization and smarter memory, and Flova 1.5 becomes an interesting upgrade for everyday production. @Flovaai flova.ai/?refCode=MPLHELP7
🤝 Paid partnership
Introducing Flova 1.5
Smarter. More Adaptive. Bring Ideas to Life.
✨ Run multiple sessions in parallel and explore your creativity freely
✨ Chat at any step and switch between creative tasks with ease
✨ Build and reuse Skills to keep expanding your creative capabilities
✨ Rediscover inspiration quickly with smart memory and retrieval
✨ Manage assets flexibly and keep your creative work organized
✨ Take control of your Agent and create at your own pace
🎊 Smart and flexible—turn ideas into videos, your way!
Neo retweeted
TAX LOOPHOLES THAT FEEL ILLEGAL, BUT ARE 100% LEGAL
1. RENT YOUR HOME TO YOUR OWN BUSINESS
Neo retweeted
Codex is becoming a real productivity partner, not just a tool for testing prompts. 🚀 Automating repetitive tasks like file organization, numbering, and verification can save so much time for the work that really matters!
@GamsGo_Global
🤝 Paid partnership
最近分享日常用 Codex 干活，结果不少朋友说想试，但直接卡在了海外订阅支付这一步。
我去看了眼 @GamsGo_Global 的 Codex 专区，选一次性付款可以直接用支付宝，不用折腾外币卡。
比如 GamsGo 自营的 Codex Plus 独享账号，1 个月订阅页面价 $15.49 USD（注意切到一次性付款和首页自动续费价略有不同，具体以下单当天页面为准），比 OpenAI 官方 $20/月划算不少，还能顺手勾选 GamsCare+ 全周期退款保障。
说回我最近拿 Codex 干的活！
手头在做的一个项目，光是前期准备就要处理几十个零散素材：拆解清单、统一文件编号、格式校验，还要核对界面排版。
我直接让 Codex 帮我把这些琐碎活接过去：用代码批量对齐文件编号，改过一版素材后自动对回原编号，再回头检查哪一处漏了状态。
有一轮检查它直接帮我揪出两处编号错位和时长对不上的细节 😂 虽然核心创意和最终成片还得自己盯，但这种整理文件、来回核对的苦力活交给 Codex，效率真的高出一截。
所以比起单纯拿它跑测试题，我更推荐直接拿手头正在做的项目去试。
卡在支付这一步的朋友可以看看 @GamsGo_Global 这个入口，用优惠码 DVDPM 首单可再减 3%（折后 $15.03/月），链接放第一条回复 👇