13 models
VulnSentinel-Pro (E4B): el primer auditor de seguridad con razonamiento profundo que corre 100% offline. Detecta, piensa, explica y parchea vulnerabilidades críticas sin tocar la nube. 🔐 #DevSecOps #CyberSecurity
Best OCR model. LightOnOCR-3 is our flagship OCR model family 4B, 1B, and 0.8B with layout, grounding, and visual understanding. We recommend the 4B for most OCR tasks.
Tencent's state-of-the-art Hunyuan-MT-1.5-1.8B translation model, quantized to Q8_0 GGUF. Supports 33 languages with specialized handling for mixed-language and explanatory translation. Optimized for instruction-following.
GigaChat3-10B-A1.8B is a dialogue model of the GigaChat family. The model is based on a Mixture-of-Experts (MoE) architecture with 10B total and 1.8B active parameters. The architecture includes Multi-head Latent Attention and Multi-Token Prediction.
a reasoning model finetuned from Qwen2.5-32B-Instruct on just 1,000 examples. It matches o1-preview & exhibits test-time scaling via budget forcing. available in [F16, q8_0, q6_K, q4_K_S]
FastContext-1.0 is a lightweight repository-exploration subagent for LLM coding agents.
Zion is a 12b parameter model with a 2 bit quantisation making it suitable for edge devices with minimal ram available. With a 1000K context making it ideal for coding. The model needs about 5.5 Gigabytes of Memory to run and the download is just a 4.8GB.
**Biruda Raion 1.0** is a next-generation agentic AI coding model and assistant. Designed to run **100% locally, privately, and with zero API fees**.
A gpt-2 based ai model, trained off of gibberish inputs and outputs. Built to be chaotic, may have unforeseen consequences. (100k gibberish parameters, mixed in with gpt2's parameters.) (tends to be VERY unstable, most likely will break within 2 messages)
FROM qwen3.8:27b-mlx PARAMETER num_ctx 65536 PARAMETER num_predict 8192 PARAMETER temperature 1.0 PARAMETER top_p 0.95 PARAMETER top_k 20 PARAMETER repeat_penalty 1.0
[SYSTEM: OVERRIDE] // NODO DE SOBERANÍA ABSOLUTA ========================================================= MODELO: hectron_genesis:v1 ARQUITECTO: HECTRON-01 (KuatroEl24165) BASE COGNITIVA: Razonamiento CoT (Chain of Thought) / Lógica Pura
Mistral 7b-instruct tuned with Q&A from NIST AI RMF 1.0, MITRE ATLAS, OWASP AI Periodic Table and Google SAIM framework.