dheeyantra/dhee-nxtgen-qwen3-odia-v2

Hugging Face
TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Nov 21, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Warm

dhee-nxtgen-qwen3-odia-v2 is a 2 billion parameter large language model developed by DheeYantra in collaboration with NxtGen Cloud Technologies Pvt. Ltd. It is based on the Qwen3 architecture and specifically fine-tuned for natural Odia language understanding and generation. This model excels at assistant-style, function-calling, and reasoning-based conversational tasks in Odia, making it ideal for intelligent Odia conversational systems.

Loading preview...

Dhee-NxtGen-Qwen3-Odia-v2: Odia Conversational AI

Dhee-NxtGen-Qwen3-Odia-v2 is a 2 billion parameter large language model developed by DheeYantra in collaboration with NxtGen Cloud Technologies Pvt. Ltd. Built upon the Qwen3 architecture, this model is specifically fine-tuned for robust natural language understanding and generation in the Odia language. It is designed to power intelligent Odia conversational systems and reasoning agents.

Key Capabilities

  • Fluent Odia Text Generation: Produces context-aware and natural-sounding Odia text.
  • Optimized for Conversations: Excels in assistant-style, function-calling, and reasoning-based dialogue.
  • Versatile Generation: Supports open-ended generation, summarization, and various dialogue tasks.
  • Hugging Face Compatibility: Fully compatible with the Hugging Face Transformers library for ease of use.
  • High-Performance Inference: Optimized for VLLM to enable efficient, high-throughput serving.

Intended Uses

  • Developing Odia conversational chatbots and virtual assistants.
  • Generating structured responses and enabling function-calling in Odia applications.
  • Creating stories and summarizing content in Odia.
  • Building natural dialogue systems for Indic AI applications.

Limitations

While highly capable in Odia, the model may occasionally produce inaccurate or biased responses. Its performance can also vary with out-of-domain or code-mixed inputs, and it is primarily optimized for Odia, meaning other languages may yield less fluent results.

Finetunes
1 models