Distilled models for Kimi-2.5, Gemini 3.1 Pro
AI & ML interests
Optimising LLM/SLMs for real-world impact.
Recent Activity
View all activity
Organization Card
We specialize in fine-tuning and optimizing Large and Small Language Models (LLMs/SLMs).
KhazarAI is a community-driven initiative dedicated to developing open-source language models and datasets. Our mission is to support domain-specific NLP research and enable organizations to leverage AI for smarter decision-making, process automation, and data-driven insights.
We focus on:
- Training and publishing instruction-tuned and reasoning language models using SFT, GRPO, and RLHF for the different domains
- Providing tools and resources for researchers, developers, and analysts
- Enhancing the adoption of AI in business processes and decision-making
Join us in shaping the future of AI-driven solutions.
models 32
khazarai/Math-VL-8B
Image-Text-to-Text • 9B • Updated • 83 • 1
khazarai/Bio-8B-it
Text Generation • 8B • Updated • 989 • 1
khazarai/Qwen3-4B-Kimi2.5-Reasoning-Distilled-GGUF
Text Generation • 4B • Updated • 3.61k • 3
khazarai/Qwen3-4B-Gemini-3.1-Pro-Reasoning-Distilled
Text Generation • 4B • Updated • 5.1k • 2
khazarai/Qwen3-4B-Kimi2.5-Reasoning-Distilled
Text Generation • 4B • Updated • 861 • 1
khazarai/Quantum-ToT
Text Generation • 2B • Updated • 216 • 1
khazarai/HeisenbergQ-0.5B-RL
Text Generation • 0.5B • Updated • 426 • 1
khazarai/Medical-QA
Text Generation • 0.6B • Updated • 420 • 1
khazarai/BioGenesis-ToT
Text Generation • 2B • Updated • 218 • 1
khazarai/Personal-Finance-R2
Text Generation • 2B • Updated • 418 • 1