Chirp Chirp! 馃惁 We are introducing Ornith-1.5, a major step toward building foundation models through end-to-end self-improvement.
DeepSeek-OCR is a vision-language model that can perform token-efficient OCR.
Beam is an open model by Reflection designed to deliver strong performance for areas such as agentic tool use, coding and general reasoning with an emphasis on token efficient usage.
The TinyLlama project is an open endeavor to train a compact 1.1B Llama model on 3 trillion tokens.
A series of models from Groq that represent a significant advancement in open-source AI capabilities for tool use/function calling.
An 8B tool-calling model that picks the right tool with valid arguments and does not loop. Fine-tuned from Ministral 3 8B on production agent traffic. Apache-2.0.
OLMo 2 is a new family of 7B and 13B models trained on up to 5T tokens. These models are on par with or better than equivalently sized fully open models, and competitive with open-weight models such as Llama 3.1 on English academic benchmarks.
pulled from https://huggingface.co/second-state/Octopus-v2-GGUF/blob/main/Octopus-v2-Q4_K_M.gguf
A coding assistant with a 32k context window.
Muse Spark Open is a natively multimodal model with 816B total parameters (42B active) and a 1M-token context window. Built for coding and agentic work, it excels at long-horizon tasks, complex software engineering, and multi-step reasoning.
NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts (MoE) model with 3B active parameters built for always-on agents.
Laguna XS 2.1 is a 33B total parameter Mixture-of-Experts model with 3B activated parameters per token designed for agentic coding and long-horizon work on a local machine.
GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder鈥揹ecoder architecture.
Qwen3.5 0.8B in Q8_0 quantization. Small, fast model with 262K context, vision, tool use, and thinking capabilities. Optimized for local/edge deployment on constrained hardware.
A balanced configuration variant of Qwen3.5-4B, planned for FIM (Fill-In-the-Middle) and Tool Calling in restrict capacity environments. Using the Q3_K_S GGUF quantization from HuggingFace (unsloth).