beyoru/Tama-JP-beta

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Sep 30, 2025License:mitArchitecture:Transformer0.0K Open Weights Warm

Tama-JP-beta is a 4 billion parameter conversational AI model, continuously pretrained on Qwen3-4B, developed by Beyoru. It is specifically optimized for immersive NSFW roleplay and natural chatting, distinguishing itself through its capability to generate explicit content. With a notable context length of 40960 tokens, this model is designed for use cases requiring detailed and extended roleplay-style conversations.

Loading preview...

Tama-JP-beta: Conversational AI for Immersive Roleplay

Tama-JP-beta is a 4 billion parameter conversational AI model, developed by Beyoru, that has been continuously pretrained on the Qwen3-4B architecture. Its primary distinction lies in its specialized optimization for immersive NSFW roleplay (RP) and natural chatting experiences. The model is designed to handle detailed and extended conversational flows, supported by a substantial context length of 40960 tokens.

Key Capabilities

  • Specialized for NSFW Roleplay: Engineered to generate explicit content and engage in immersive roleplay scenarios.
  • Natural Chatting: Beyond roleplay, it supports general conversational interactions.
  • Extended Context: Features a 40960-token context window, allowing for longer and more coherent dialogues.

Good For

  • Creating detailed character interactions: Ideal for scenarios where users define character prompts for rich, interactive storytelling.
  • Applications requiring explicit content generation: Suitable for use cases that specifically permit or require NSFW outputs.
  • Long-form conversational agents: Its large context window makes it effective for maintaining continuity over extended chats.