CoolFace
Modelpublic

nex-agi/DeepSeek-V3.1-Nex-N1.1

sourceHugging Faceapache-2.0updated 8mo agoView on Hugging Face
5likes54downloads
Model Card

<div align="center"> <img src="./figures/NEX_logo.svg" width="20%"/> </div>

<div align="center"> 🏠 <a href="https://nex.sii.edu.cn"><b>Home&nbspPage</b></a>&nbsp&nbsp | &nbsp&nbsp 🤗 <a href="https://hf.co/collections/nex-agi/nex-n1"><b>Model</b></a>&nbsp&nbsp | &nbsp&nbsp 🤗 <a href="https://huggingface.co/datasets/nex-agi/agent-sft"><b>Data</b></a>&nbsp&nbsp | &nbsp&nbsp 📚 <a href="https://huggingface.co/papers/2512.04987"><b>Paper</b></a>&nbsp&nbsp | &nbsp&nbsp 💻 <a href="https://github.com/nex-agi/Nex-N1"><b>Code</b></a> </div>

Nex-N1.1

Nex is a next-generation, full-stack agentic platform that brings foundation models, synthetic data pipelines, RL training, agent frameworks, and deployment tools together in one unified ecosystem. DeepSeek-V3.1-Nex-N1.1 is the latest flagship release of the Nex-N1 series — a post-trained model designed to highlight agent autonomy, tool use, and real-world productivity. We are committed to making it easier than ever to build and deploy AI agents by offering researchers and entrepreneurs a high-performance, reliable, and cost-effective "out-of-the-box" agent system.

What's New in N1.1

  • —Enhanced Claude Skills Integration: Significantly improved ability to work with Claude Skills, delivering a more seamless tool-calling experience
  • —Improved Instruction Following: Substantially enhanced instruction adherence with greater stability and accuracy in complex tasks
  • —Upgraded Frontend Capabilities: Specialized optimization for frontend development scenarios, excelling in HTML, CSS, and JavaScript generation
  • —Advanced Vibe Coding: Further enhanced general-purpose coding abilities to meet diverse development needs
  • —200K Context Length: Extended context window to 200K tokens, perfectly compatible with Claude Code and other long-context scenarios
  • —MTP Support: Added Multi-Token Prediction capability for improved inference efficiency

Usage

Local Deployment

We recommend sglang for serving Nex-series models locally:

bash
python -m sglang.launch_server --model-path /path/to/your/model

Function Calling

Nex-series models support robust function-calling capabilities. To maximize the function-calling capabilities of the Nex-series models, we modified the tool parser of qwen3_coder, see: <https://github.com/sgl-project/sglang/pull/13411>. To enable this feature, simply add the --tool-call-parser qwen3_coder flag when launching the server:

bash
python -m sglang.launch_server --model-path /path/to/your/model --tool-call-parser qwen3_coder

Mini Program Development

Nex-N1.1 is optimized for mini program development with enhanced 200K context support. For optimal performance, we recommend using Claude Code configured with both context7 and a search MCP.

shell
claude mcp add --transport http context7 https://mcp.context7.com/mcp --header "CONTEXT7_API_KEY: [CONTEXT7_API_KEY]"

claude mcp add --transport stdio serper-search --env SERPER_API_KEY=[SERPER_API_KEY]  -- npx -y serper-search-scrape-mcp-server

Refer to <https://github.com/upstash/context7> for more details on setting up context7.