Nex retweeted
Meet Nex-N2.5-Max: a new conversational text-generation model with 11,404 downloads. Built on deepseek_v4 architecture and fp8 optimized. It's ready for your chat apps and creative projects. Let's explore what it can do.
Nex retweeted
omg our @NexEcosystem nex-n2.5 mini just hit #3 on @hugging face trending LLM 🤗 🏆
and PRO is right behind at #4 🥹
both can see and use your computer to get things done, and they're FREE on @OpenRouter!
let’s gooo
Nex retweeted
🚀 Free API access for Nex-N2.5-Pro and Nex-N2.5-mini is now live on ModelScope! Give it a try!
🔹 Nex-N2.5-Pro: modelscope.ai/models/nex-agi…
🔹 Nex-N2.5-mini: modelscope.ai/models/nex-agi…
⏰ Free trial runs until September 24 (Beijing Time).
Built for long-horizon agentic tasks, the Nex-N2.5 family excels at computer use, web browsing, and visually grounded workflows.
An agent model, plus two multimodal models that can see and operate computers. Nex-AGI releases Nex-N2.5 in mini, Pro, and Max sizes.📜 Apache 2.0.
🤖modelscope.ai/collections/ne…
🏆 Max leads the published BrowseComp comparison with 92.6. Pro ranks first on OSWorld-G with 87.4.
🖥️ The 35B mini and 397B Pro browse the web, operate software, verify results, and self-correct through visual feedback.
🧠 The 1.6T Max targets complex text-based agent tasks, from coding and tool use to research and knowledge work. It marks Nex-AGI’s first full post-training run at trillion-parameter scale.
⚙️ All three support function calling and direct, adaptive, or full thinking modes.
Nex retweeted
What happens when an AI stops telling you how to build something and actually builds it?
Nex-N2.5 Pro took an idea from concept to playable game:
→ Rough concept
→ 2D design
→ Blender 3D environment
→ Working game
The interesting part wasn’t the final result.
It was the process.
The model interacted directly with the tools, observed the changes on screen, made adjustments, and continued working instead of simply generating instructions.
That’s the real computer-use test.
Can an AI:
• Open Blender and build a scene?
• Inspect the output and iterate?
• Write a script, tweak parameters, and troubleshoot?
• Hit a limitation, switch to the GUI, and finish the task?
Nex-N2.5 Pro is starting to make that workflow feel surprisingly real.
You can try it on OpenRouter for a limited time:
Try Nex-N2.5 Pro on OpenRouter
#NexN2 #Blender #ComputerUse #OpenRouter
🤝 Paid partnership
Nex retweeted
Nex AGI's agentic N2.5 family is live on OpenRouter: Nex-N2.5-Pro and Nex-N2.5-Mini.
Both are built to turn goals into working, verified outcomes, with agentic coding as the core strength. Tool calling and adjustable reasoning effort on both.
Free for a limited time ↓
openrouter.ai/nex-agi
💥 CAD Agent 又往前走了一步!Nex-N2.5 Pro 从零建模机械臂,最后直接交付 SolidWorks 工程文件。
整个流程不是生成一张 3D 效果图,而是全程调用 CAD 工具完成设计,你可以看着它一步一步操作你的电脑:
✅ 拆分底座、关节、连杆和夹爪
✅ 生成并执行 CAD 自动化脚本
✅ 遇到 BLOCKER 后修改方案继续跑
✅ 完成零件建模、整体装配和爆炸视图
✅ 输出 robot_arm.SLDASM、运行记录与验证结果
从一句需求,到一套可以打开、拆解和继续修改的 CAD 项目。
这个完成度确实有点夸张了 🤖
这将使得未来机械设计变得更加高效,快速
另外最重要的是Nex-N2.5 Pro 正在 OpenRouter 限时免费开放:
openrouter.ai/nex-agi
#NexN25 #CAD #ComputerUse #AIAgentControl
Nex retweeted
i love GPT-6 Astra, so when i heard we were releasing an agentic computer-use model, i got VERY excited.
i kept asking the team: can Nex-N2.5 Pro do that thing like GPT-6 launch video? go from a 2D design into 3D modeling in Blender?
turns out, yes 😭
and then it went a step further and made an interactive HTML game!!
i keep thinking about what comes after this. steam power helped make railways possible. short-form video brought us TikTok, so now we scroll infinitely and shop in the same app.
really wonder what comes next when AI can actually use the tools we use and do it 10x better than me? 😂
Meet Nex-N2.5 family, our latest open source agentic models.
Mini (35B) and Pro (397B) bring multimodal understanding and computer use. Max (1.6T) is a text-only MoE model built for complex reasoning, coding, and agent workflows.
Strong benchmark results in workflow automation and computer use:
⚡ Max scores 50.2 on AutomationBench v1.0.6—just 0.1 points behind Claude Opus 5
⚡ Pro scores 56.4 on OSWorld-2, ahead of Qwen3.8-Max at 46.7
With continuous action, visual feedback, and self-correction, NEX-N2.5 can work across Blender and CAD tools, handle expense reimbursements, and even play PC games—turning visual understanding into sustained, real-world execution.
🔗@huggingface
huggingface.co/nex-agi/Nex-N… huggingface.co/nex-agi/Nex-N… huggingface.co/nex-agi/Nex-N…
🔗@modelscope
modelscope.cn/models/nex-agi… modelscope.cn/models/nex-agi… modelscope.cn/models/nex-agi…
🔗 Github github.com/nex-agi/Nex-N2.5
🔗Website nex-agi.com/
Nex retweeted
I asked Claude to help me verify the claim:
------
I (Claude) independently verified the claim that Rio-3.5-Open-397B is a weight merge of Nex and Qwen. It checks out.
A developer opened an issue claiming that prefeitura-rio/Rio-3.5-Open-397B is just a ~0.6/0.4 linear blend of the Nex-N2-Pro model and the official Qwen3.5-397B-A17B base, with no original training.
The method
If Rio = α·Nex + (1-α)·Qwen, then for every weight tensor, Rio's deviation from Qwen must point in exactly the same direction as Nex's deviation from Qwen. Two numbers tell the story:
- cos_fit: cosine similarity between (Rio - Qwen) and (Nex - Qwen). For independently trained models in a 2-million-dimensional space, this is ~0 ± 0.0007. For a merge, it's ~1.
- α: how far Rio sits along the line from Qwen toward Nex.
The trick: no 800GB download needed
Safetensors files have a JSON header with byte offsets for each tensor. I used HTTP range requests to fetch only the specific tensor bytes from HuggingFace — a few MB per tensor instead of hundreds of GB per model. Entire verification runs on a laptop.
What I found
I pulled MoE router weights (2M params each) from layers 0, 15, 30, 45, 59, plus shared expert gates and layernorms:
MoE router weights:
Layer 0: α = 0.573, cos_fit = 0.992
Layer 15: α = 0.647, cos_fit = 0.962
Layer 30: α = 0.627, cos_fit = 0.967
Layer 45: α = 0.582, cos_fit = 0.987
Layer 59: α = 0.567, cos_fit = 0.997
Shared expert gates:
Layer 0: α = 0.568, cos_fit = 0.997
Layer 30: α = 0.581, cos_fit = 0.988
What this means
A cos_fit of 0.99 in a 2-million-dimensional space is not "high similarity." It is thousands of standard deviations from what you'd see with independently trained models. There is no innocent explanation.
The recovered α clusters tightly around 0.57 across all layers — matching nex-agi's claim of 0.571 almost exactly. This is one model poured into another at a fixed ratio.
(Layernorm weights show a higher α ~0.9. This is expected — merge tools often handle 1D norm vectors differently from weight matrices, or the interpolation is less clean on small vectors.)
Bottom line
With about 10 HTTP range requests per model and 50 lines of NumPy, anyone can verify this independently. The math is unambiguous: Rio-3.5-Open-397B is approximately 57% Nex-N2-Pro + 43% Qwen3.5-397B-A17B.
Code that you can run for yourself: gist.github.com/xianbaoqian/…
Nex retweeted
wait... what???
github.com/nex-agi/Nex-N2/is…
Nex retweeted
Nex-N2 is now open source!An agentic model series from Nex AGI built for coding, tool use, deep research, and long-horizon workflows. 🧠🔎
🛠️ modelscope.ai/models/nex-agi…
⚙️ modelscope.ai/models/nex-agi…
● Models: Nex-N2-Pro 397B total, 17B active; Nex-N2-mini 35B total, 3B active
● Agentic Thinking: adaptive reasoning depth + coherent reasoning across coding, search, tool calling, and execution
● Efficiency: Nex-N2-mini saves roughly 20% overall token cost vs forced thinking while matching or slightly exceeding task performance
● Open-model lead: 75.3 on Terminal-Bench 2.1, 80.8 on SWE-Bench Verified, 83.7 on BrowseComp, and 1585 on GDPval among listed open baselines
● Deployment: customized SGLang fork, reasoning parser, tool-call parser, Docker image
● License: Apache 2.0
Nex retweeted
🚀 Nex-N2-Pro from @NexEcosystem is now available on Novita AI.
An open-source agentic reasoning model post-trained on Qwen3.5-397B-A17B MoE, built for coding agents, software engineering, and deep research workflows.
Nex-N2-Pro brings:
• Agentic Thinking for complex workflows
Unifies reasoning, tool use, and environment execution for long-horizon tasks
• Strong coding + terminal performance
Scores 75.3 on Terminal-Bench 2.1 and 80.8 on SWE-Bench Verified
• Designed for self-evolving harnesses
Specifically optimized for Agentic Harness Engineering (AHE), with top-tier pass@1 performance reaching 69.0% in a SQL interpreter self-evolution task
• Fast, developer-ready access
Run Nex-N2-Pro through Novita’s API with simple integration
Excited to bring Nex-N2-Pro to developers on Novita.
Nex retweeted
Introduce NexRL: Extreme Usability
We aim to develop an extremely user-friendly RL framework, with four main principles:
1. Extremely Lightweight
2. Extremely Community-Driven
3. Extremely Extensible
4. Long-Term Maintainability
NexRL adopts an ultra-loosely-coupled architectural design, with core characteristics embodied in component modularization and training-inference service-orientation.
Along with Nex ecosystem @NexEcosystem , NexRL has already boosted agent training, on-policy-distillation, and training API serving (see our previous blog dawning-road.github.io/blog/…).
Details at: dawning-road.github.io/blog/…
Nex retweeted
Introduce Scalable Interactive Oversight.
In vibe coding, users often fall into prolonged feedback–iteration loops. This reflects a scalable oversight problem: humans act as "weak supervisors", struggling to precisely articulate complex intent and to verify long-horizon outputs; meanwhile, models function as "strong executors", capable of efficient task execution but increasingly difficult to steer with reliable alignment signals.
Scalable Interactive Oversight decomposes complex intent into a recursive tree of manageable decisions to amplify human supervision. It can be optimized via Reinforcement Learning using only online user feedback, offering a practical pathway for maintaining human control as AI scales.
Details at: dawning-road.github.io/blog/…
Nex retweeted
We release ABC-bench with @Open_MOSS. It benchmarks the agents' ability to reliably deliver production-grade backend services.
Findings:
1. Full-lifecycle tasks pose significant challenges for all models.
2. Models generally lack cross-language robustness.
3. Environment configuration is the primary source of performance bottlenecks.
Details at: dawning-road.github.io/blog/…
Nex retweeted
Introducing Weaver: our Training-as-a-Service platform.
Weaver is a training API designed for researchers and developers working with LLMs, inspired by Tinker @thinkymachines . As an ecosystem-oriented product, it seamlessly integrates with agent frameworks and reinforcement learning training frameworks, enabling flexible fine-tuning of agentic models. Weaver lets you focus on what matters – your data, algorithms and agents – while handling the complexity of distributed training infrastructure.
Focus on the science. We’ll handle the rest.
Details at: dawning-road.github.io/blog/…
Nex retweeted
FP8 series Part II:
We introduce a ZeRO optimizer implementation scheme adapted for Hopper architecture and oriented toward Blockwise quantization.
This optimizer has been thoroughly validated in our team's internal native FP8 SFT and RL tasks based on the DeepSeek architecture.
Details at: dawning-road.github.io/blog/…
Nex retweeted
On-policy RL is stable but slow. Off-policy RL is efficient but unstable.
Our new analysis reveals why off-policy breaks—optimization imbalance + entropy clipping effects.
Meet BAPO: a balanced, adaptive clipping method that finally makes off-policy RL stable for long-horizon LLM training.
Details at: dawning-road.github.io/blog/…
Nex retweeted
Meet Nex by NEX-AGI — a non-thinking model built for agents that crushes it in coding, tool use, and roleplay 🚀
✅ SOTA among open models on Tau2-Bench, BFCL V4, GAIA2
✅ Top-tier in frontend, vibe coding, and mini-program/backend dev (human eval confirmed)
✅ Plug-and-play with Claude Code, Cursor, etc.
👉 Expolore Nex: modelscope.cn/organization/n…
🔥 Free for now on SiliconFlow: siliconflow.com/models/deeps…
Nex retweeted
Welcome Nex-N1, a new series of agentic foundational models, to @huggingface
- available in different sizes from 8B, 30B, 32B to 671B
- strong in tool-use, web-search and real-world agentic workflow
- some SFT dataset has been open sourced
Technical report come up soon!
Nex retweeted
We rebuilt Grouped Per-Token Quantization at the kernel level — removing the real bottleneck in FP8/MoE training and pushing efficiency to the hardware limit:
• 20× faster core ops
• ~2.7 TB/s effective throughput
Details at: dawning-road.github.io/blog/…





























