Our first generation of open-weight models
AI & ML interests
Open-source reasoning models, multimodal systems, and private mobile-first local AI.
Recent Activity
Organization Card
rrchat AI
Open-source compact reasoning models & private mobile-ready local AI
🌐 Overview
rrchat AI is an open-source initiative focused on high-efficiency, private local AI: combining compact bilingual (RU/EN) reasoning models with native mobile interfaces.
Our goal is to bring sovereign intelligence directly into your workflow — zero cloud telemetry, maximum throughput on consumer hardware, and seamless orchestration between mobile devices and local GPU workstations.
🧠 Models: RR-Gen1 Series
The RR-Gen1 lineup spans from ultra-fast 1.7B edge assistants to a 9B agentic flagship, optimized for agentic tool use, mathematical reasoning, and competitive coding:
| Model | Size / Active | Base Architecture | Primary Strengths | Recommended VRAM | Weights & GGUF |
|---|---|---|---|---|---|
| Viper-9B (Flagship) | 9B | Qwen3.5-9B | Competitive Coding (LCB 33%), Agentic Tool Calling, Math CoT, R-Tuning Honesty | ~8–12 GB | Base · GGUF |
| Cerberus-E4B | 4B (7.5B Total) | Gemma 4-E4B | Balanced Multi-Task Reasoning, Native Gemma Tool Schemas, Multimodal | ~4–8 GB | Base · GGUF |
| Mercury-1.7B | 1.7B | LFM3-VL-2B | Extreme Speed (~180–200 t/s), Linear Attention + Short-Conv, SigLIP Vision | ~2–4 GB | Base · GGUF |
| Mercury-Acad-1.7B | 1.7B | LFM3-VL-2B | Academic & Scientific Reasoning, Mathematical CoT, STEM Domain Specialist | ~2–4 GB | Base · GGUF |
🔬 Lineup Highlights
- 🐍 Viper-9B: The flagship for software engineering and complex reasoning. Demonstrates +14% absolute gain on LiveCodeBench v6, 84.7% on MBPP+, and calibrated factual honesty without confident hallucinations.
- 🐺 Cerberus-E4B: Efficient mid-weight specialist built on Google's Gemma 4 backbone, excelling in multi-turn tool interaction and structured bilingual reasoning.
- ⚡ Mercury-1.7B: Sub-2B powerhouse powered by Liquid AI's hybrid linear attention and causal depthwise convolutions, delivering desktop-class reasoning at up to 200 tokens/sec on edge hardware.
- 🎓 Mercury-Acad-1.7B: Tailored for mathematical research, paper comprehension, and STEM problem solving with dual-mode explicit reasoning (
<think>).
📱 Client: rrchat
rrchat is a privacy-first mobile client designed to pair your phone directly with your PC or local server running llama.cpp, llama-server, or Ollama:
- 🔒 100% Local & Sovereign: Direct peer-to-peer connection over your local network/VPN. No middleman servers, no tracking, zero cloud lock-in.
- 👁️ Full Multimodal Workflow: Integrated camera vision, audio voice notes transcription, and document analysis.
- 🌐 Agentic Web Search: Privacy-preserving web exploration powered by local models without vendor profiling.
- ⚡ Instant Pairing: One-tap connect via local IP / QR code with automatic model detection and context management.
Built by rrchat · Empowering sovereign open-source intelligence
models 8
rrchat/Viper-9B-GGUF
Image-Text-to-Text • 9B • Updated • 312 • 1
rrchat/Viper-9B
Image-Text-to-Text • 9B • Updated • 26 • 1
rrchat/Mercury-Acad-1.7B
Image-Text-to-Text • 2B • Updated • 40 • 1
rrchat/Mercury-Acad-1.7B-GGUF
Image-Text-to-Text • 2B • Updated • 274 • 1
rrchat/Mercury-1.7B-GGUF
Image-Text-to-Text • 2B • Updated • 345 • 1
rrchat/Cerberus-E4B-GGUF
Image-Text-to-Text • 8B • Updated • 937 • 3
rrchat/Mercury-1.7B
Image-Text-to-Text • 2B • Updated • 36 • 2
rrchat/Cerberus-E4B
Text Generation • Updated • 22 • 2
datasets 0
None public yet