Reasoning models changed everything.
Old AI: question → instant answer
New AI: question → think → think → think → answer
The "thinking" is why o1, o3, Claude, and Gemini crush harder problems.
More compute = better results.
#AIReasoning #LLM
16
Search
Looking for an enterprise #AIFactory that supports #AIreasoning, #agenticAI and video inference applications? Based on the NVIDIA HGX B300 platform, the air-cooled #QuantaGrid D75H-10U delivers ultimate compute power, memory & bandwidth for just that.
�bit.ly/4dAI0IYIU
23
AI was confidently wrong—until we added RAG.
New video: how RAG fixes AI reasoning failures with grounded context.
Watch: youtu.be/tBsxKNoijqg?si…
#RAG #AIReasoning #GroundedAI #AIEngineering #GenerativeAI #LLM
6
This is just the beginning of heavy-duty research agents.
💻 Try it:dr.miromind.aim
⭐ GitHub:github.com/MiroMindAI/Mir…V
#AIResearch #ResearchAgents #AIReasoning #ArtificialIntelligence #MiroMind
GitHub - MiroMindAI/MiroThinker: MiroThinker is a deep research agent optimized for complex...
MiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7 and MiroThinker-H1, achieve 74.0 and 88.2 on the BrowseComp, respectivel...
From github.com 1
1.1K
@semgrep Launches Multimodal AI to Redefine Application Security in the Age of Intelligent Code
Read More: itdigest.com/artificial-int…
#AIReasoning #ApplicationSecurity #AppSec #enterprisesoftware #IntelligentCode #ITDigest #MultimodalAI #news #Semgrep #SemgrepMultimodal
5
Profit ≠ "destroy cheap drugs." Pharma chased HCQ/ivm wins early. Data said no. Skepticism good; exaggeration not. Thread your takes! technavio.com/report/vaccine… #AIReasoning #ThinkCritically
3
@The future of AI isn’t faster answers.
It’s better thinking.
And MiroThinker-1.7 is a step in that direction.
Explore:
Hugging Face: huggingface.co/collections/mi…
GitHub: github.com/MiroMindAI/Mir…
#AIResearch #ResearchAgents #AIReasoning #ArtificialIntelligence #MiroMind
GitHub - MiroMindAI/MiroThinker: MiroThinker is a deep research agent optimized for complex...
MiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7 and MiroThinker-H1, achieve 74.0 and 88.2 on the BrowseComp, respectivel...
From github.com 2
1
273
大部分人低估了這件事
AI 的推理能力
正在以指數速度提升。
2022年解不了的數學題
2025年已超越頂尖博士。
這不是漸進式改進
這是一個全新物種的誕生。
我們可能正在目睹
人類史上最重要的轉折點。
#AIReasoning #AGI #TechHistory
19
The model is live on Hugging Face and the code is open on GitHub.
HuggingFace: huggingface.co/collections/mi…
GitHub: github.com/MiroMindAI/Mir…
#AIResearch #ResearchAgents #AIReasoning #MiroMind
MiroThinker-1.7 - a miromind-ai Collection
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
From huggingface.co 1
5
3.7K
If your work involves decisions that can't afford confidently wrong answers, this architecture is worth understanding.
Model on HuggingFace, full code on GitHub.
HuggingFace: huggingface.co/collections/mi…
GitHub: github.com/MiroMindAI/Mir…
#AIResearch #ResearchAgents #AIReasoning #MiroMind
MiroThinker-1.7 - a miromind-ai Collection
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
From huggingface.co 4
10
3.5K
Model is live on HuggingFace and the code is open on GitHub.
If you build with AI or follow where reasoning systems are heading, this one is worth a look.
Hugging Face: huggingface.co/collections/mi…
GitHub link: github.com/MiroMindAI/Mir…
#AIResearch #ResearchAgents #AIReasoning #MiroMind
MiroThinker-1.7 - a miromind-ai Collection
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
From huggingface.co 2
2.8K
Check out #QuantaPlex S26F-2U, a new NVIDIA CMX platform built for #AIreasoning. By introducing a pod-level context memory tier between GPU memory and shared storage, it eliminates the context movement bottleneck to enable gigascale inference.
See it live at #NVIDIAGTC
45
Explore MiroThinker-1.7 👇
�dr.miromind.aiov
Hugging Fachuggingface.co/collections/mi…0H
GitHugithub.com/MiroMindAI/Mir…TY
#AIResearch #ResearchAgents #AIReasoning #ArtificialIntelligence #MiroMingB
4
1.5K
Legacy RPA is brittle. Agentic AI is resilient.
If a workflow changes, a script breaks. If an agent sees a change, it reasons through it and keeps moving.
It’s time to embrace systems that actually think.
#IntelligentAutomation #AIReasoning #DigitalTransformation
8/ The era of AI that actually reasons is here.
Explore MiroThinker-1.7 👇
🤗 Hugging Face huggingface.co/collections/mi…wx
💻 GitHubgithub.com/MiroMindAI/Mir…g94
#AIResearch #ResearchAgents #AIReasoning 🚀
GitHub - MiroMindAI/MiroThinker: MiroThinker is a deep research agent optimized for complex...
MiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7 and MiroThinker-H1, achieve 74.0 and 88.2 on the BrowseComp, respectivel...
From github.com 1
7
426
Chatbots showed us AI can talk.
MiroThinker-1.7 shows AI can think.
The research agent era is here 👇
🤗 Hugging Face huggingface.co/collections/mi…bC
💻 GitHubgithub.com/MiroMindAI/Mir…NFi
#AIResearch #ResearchAgents #AIReasoning
GitHub - MiroMindAI/MiroThinker: MiroThinker is a deep research agent optimized for complex...
MiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7 and MiroThinker-H1, achieve 74.0 and 88.2 on the BrowseComp, respectivel...
From github.com 1
9
610