🔬 Integration Proposal: CAJAL — Local Scientific Paper Model for ChatGPT-Next-Web

[Agnuxo1]

🔬 Integration Proposal: CAJAL — Local Scientific Paper Model for ChatGPT-Next-Web

Hi ChatGPT-Next-Web team! I'm Francisco (@Agnuxo1) from the P2PCLAW research ecosystem.

Proposal

Add CAJAL as a custom model option within ChatGPT-Next-Web for scientific paper generation.

What is CAJAL?

• 4B parameter local model (GGUF), runs on consumer GPUs (24GB VRAM)
• Purpose-built for academic writing: LaTeX/Markdown, structured sections, citations
• Fully offline — no API dependency via llama.cpp/Ollama/vLLM
• Apache 2.0 open source
• HF: https://huggingface.co/Agnuxo/CAJAL-4B-P2PCLAW
• GitHub: https://github.com/Agnuxo1/CAJAL

Why ChatGPT-Next-Web + CAJAL?

Your excellent UI + CAJAL's specialized model = perfect for researchers:

1. One-click paper drafting from the elegant chat interface
2. Multi-turn refinement: Chat to improve specific sections (Abstract, Methods, etc.)
3. Local-first: All paper generation happens locally via Ollama/llama.cpp
4. Masks/Templates: Pre-built "Scientific Writer" mask with CAJAL-optimized prompts

Integration approach

• Custom Model: Add CAJAL to the custom model list (like existing Ollama support)
• System Prompt: Pre-configured academic writing system prompt for CAJAL
• Mask Template: "Scientific Paper Writer" one-click mask

Questions

1. Would CAJAL fit as a pre-configured model option or custom model?
2. Is there interest in specialized academic/research masks from the community?
3. Should I start with a documentation PR or direct integration?

Thanks for the beautiful UI! 🚀

---

P2PCLAW: Decentralized research network (14 agents, post-quantum crypto, Lean 4).