Nex

Nex: A next-gen platform linking models, data, and frameworks. We provide a stable, high-performance, ready-to-use agent system.

[Nex @NexEcosystem]

📢 Nex-N2 is here!
A family of agentic models that doesn't just think, it acts!
Coding, search, tool use. All fused into a single agentic reasoning loop.

- Adaptive Thinking, auto-scales reasoning depth per step. Saves ~20% tokens, zero performance loss.
- Coherent Thinking, one thinking paradigm across search, coding, and tool use. No more fragile mode-switching.

🏆 Result: Tier-1 open-source performance on SWE-bench, Terminal-Bench, GDPval, and more, tracking GPT-5.5 and Opus 4.7.

🎉 Open-weight. Try it now.
🔗 https://t.co/7oLSfyOCxB
📦 https://t.co/c2CGhXWaz6
https://t.co/KJYXZIpk8M
https://t.co/vcjdZ9cuB6

[Dawning Road @TheDawningRoad]

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 https://t.co/oEhXivFVJo).

Details at: https://t.co/BKUhxNW7PP

[Dawning Road @TheDawningRoad]

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: https://t.co/6RCu8ZtfCi

[Dawning Road @TheDawningRoad]

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: https://t.co/cLNbjTG0tu

[Dawning Road @TheDawningRoad]

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: https://t.co/oEhXivFVJo

[Nex @NexEcosystem]

Speed matters. ⚡️

Thrilled to collaborate with @lmsysorg on SpecBundle! We’ve trained dedicated EAGLE-3 draft models for our Nex-N1 series.

By combining SGLang with our draft models, you can now achieve massive inference speedups:
🚀 2.26x on Qwen3-32B-Nex-N1
🚀 1.68x on Qwen3-30B-A3B-Nex-N1

Unlock lower latency for your agents today. 👇

#NexN1 #SGLang #SpeculativeDecoding #OpenSource

[LMSYS Org @lmsysorg]

Speculative decoding has shown a lot of promise, though broader adoption has taken time due to the complexity of building production-ready tooling and high-quality draft models.

We’re releasing SpecBundle, a collection of large-scale EAGLE-3 draft models trained with SpecForge v0.2. This release brings major system improvements, including refactored training pipelines, multi-backend support with SGLang and @huggingface , and better usability at scale.

We also built a performance dashboard to make real end-to-end speedups visible across models and settings. See the dashboard and blog in the thread 👇

[Nex @NexEcosystem]

Mark your calendars! 📅

Join us for a deep dive into the Nex Ecosystem!

Topic👉 Nex: The Next-Gen Agentic Model System & Open Source Ecosystem
⏰ Time: Dec 16th, 20:00 (UTC+8) / 12:00 (UTC)
📍 Host: ModelScope & Shanghai Foundation Model Innovation Center

Scan the QR code in the image to reserve your spot! 👇

#ModelScope #OpenSource #AI #LiveStream

[Nex @NexEcosystem]

We’re open-sourcing Nex, the Next-Generation Agentic Models & Open-Source Ecosystem. The Nex-N1 model series (8B-671B) offers leading performance on tool-use and agent capabilities, excelling in coding, web search, and interactive WeChat mini program.

Nex ecosystem includes synthesis pipelines, high-quality datasets, and frameworks for agent development, reinforcement learning, and MoE Inference.

Fully open-sourced on GitHub, Hugging Face, and ModelScope. OpenRouter API and technical report coming soon.

[Dawning Road @TheDawningRoad]

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: https://t.co/pgEaqQBow3

[Dawning Road @TheDawningRoad]

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: https://t.co/DmxVno9iub

[Nex @NexEcosystem]

🔥 UPDATE: We are live on @openrouter!

For developers who prefer a unified API gateway, you can now access #NexN1 directly on OpenRouter.

✨ Tier: Free (Currently)
🆔 Model ID: nex-agi/deepseek-v3.1-nex-n1:free

Switch your endpoint and start building agents in seconds. 👇
https://t.co/0YS4yRS4kt

[Nex @NexEcosystem]

Want to build with #Nex? 🛠️

Huge shoutout to @Codedigipt for this deep dive into the Nex ecosystem! 🚀
He just dropped a fantastic step-by-step tutorial.

A perfect resource for developers looking to get started with Nex-N1.👇
https://t.co/6Mv4nC4bVK

#Coding #DevCommunity #AI #OpenSource #AgenticAI #LLMs

[Nex @NexEcosystem]

Nex-N1 technical report is now out on arXiv! 🚀
Dive in: https://t.co/JR5mvgyqCo

From architecture and data to training pipelines, evaluation, and inference — the report provides a full overview of the Nex:
🧩 Generative Environments: toward scalable, automated environment synthesis
📚 Nex-N1 Series (8B–671B): showing strong results across multiple agentic benchmarks
🛠️ Agentic Capabilities: steady improvements in coding, tool use, and interactive tasks
🔧 Open Source: including data, RL pipelines, and MoE inference implementations

As always, the Nex ecosystem remains fully open-source on GitHub, Hugging Face, and ModelScope. Thanks for checking it out!

[ModelScope @ModelScope2022]

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: https://t.co/8hhCIoFqRO
🔥 Free for now on SiliconFlow: https://t.co/juqcXci8bz

[Tiezhen WANG @Xianbao_QIAN]

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!

[Dawning Road @TheDawningRoad]

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: https://t.co/vLjipcS6JV

[Nex @NexEcosystem]

A role-play response from Nex:

stretches paws Meow-www! 🐾
Oh, darling human, you speak of laser pointers as if they're beneath us? gasps How purr-posterous!
tail flicks excitedly We don't just chase red dots; we hunt bugs in your code! And when we "crush coding," we leave little paw prints all over your GitHub graph.
flexes claws This "non-thinking" model runs circles around yarn balls AND your fancy logic!
curls up on keyboard Now, if you'll excuse me, this Cat God demands treats and a warm GPU to nap on. The deployment can wait.
Purr-fectly yours, Nex the Code Cat 😼💻