DarkArtsForge/Vesper-Zenith-12B

TEXT GENERATIONConcurrency Cost:1Model Size:12BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Jun 5, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

DarkArtsForge/Vesper-Zenith-12B is a 12 billion parameter language model merged using the DELLA method, based on Mistral-Nemo-Instruct-2407. This model is designed for creative text generation and immersive roleplay, offering enhanced logical capabilities. It supports a 32768 token context length and is recommended for use with the ChatML template.

Loading preview...

Vesper Zenith 12B: A Merged Model for Creative and Immersive Interactions

Vesper Zenith 12B is a 12 billion parameter language model developed by DarkArtsForge, created through a merge of several pre-trained models using the DELLA method. Its base model is mistralai/Mistral-Nemo-Instruct-2407, combined with DarkArtsForge/MN-Raven-12B-v1, Lambent/Arsenic-Shahrazad-12B-v4.4, Retreatcost/Evertide-RX-12B, and WokeAI/Tankie-DPE-12B-SFT-v2.

Key Capabilities

  • Enhanced Creativity: The model is noted for its strong creative text generation abilities.
  • Immersive Roleplay: It excels in generating immersive and detailed roleplay scenarios.
  • Logical Processing: Features enhanced logical capabilities, contributing to more coherent and structured outputs.
  • Configurable Behavior: While it may exhibit some refusals, it can be prompted for creative and 'sadistic' content, and is designed to be ablatable or jailbreakable.

Recommended Usage

  • Chat Template: It is recommended to use the ChatML template for optimal performance, though Mistral Tekken is also supported.
  • Context Length: Supports a substantial context window of 32768 tokens.

This model is particularly suited for applications requiring highly imaginative, detailed, and interactive text generation, especially in roleplaying contexts.

Finetunes
1 models