RicardoEstep/RPBizkit-v8-12B

TEXT GENERATIONPricing:Input $1.2 / Output $4.8Concurrent Unit Cost:1Model Size:12.2BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 19, 2026Architecture:Transformer Featherless Exclusive Cold

RicardoEstep/RPBizkit-v8-12B is a 12.2 billion parameter experimental language model based on a Mistral Nemo mix, created by RicardoEstep. This model utilizes Model Stock and SLERP merge methods with Mergekit, combining several 'RP Uncensored' fine-tuned models. It is specifically designed for roleplay (RP) applications, offering a 32768-token context length and supporting a 128K context size. The model prioritizes RP core logic and substance core vocabulary for enhanced creative and narrative generation.

Loading preview...

RicardoEstep/RPBizkit-v8-12B: An Experimental RP-Focused Merge

This 12.2 billion parameter model, created by RicardoEstep, is an experimental merge combining elements of the Mistral Nemo architecture. It leverages advanced merging techniques like Model Stock and SLERP with Mergekit to integrate multiple fine-tuned "RP Uncensored" models.

Key Capabilities & Design:

  • Architecture: Built upon a Mistral Nemo base, integrating several specialized fine-tunes.
  • Merging Strategy: Employs a three-part merging process:
    • RP Core: Merges models like TheDrummer/UnslopNemo-12B-v4.1, ArliAI/Mistral-Nemo-12B-ArliAI-RPMax-v1.2, and allura-org/MN-12b-RP-Ink using the Model Stock method.
    • Substance Core: Combines models such as SicariusSicariiStuff/Impish_Bloodmoon_12B, ChaoticNeutrals/Nera_Noctis-12B, and allura-org/Bigger-Body-12b also via Model Stock.
    • Final Mix: Uses SLERP to blend the RP Core and Substance Core, with specific weighting to prioritize RP Core logic and Substance Core vocabulary.
  • Context Length: Features a clean tokenizer and embedding sizes (131072), designed to support a full 128K context size.
  • Chat Template: Recommended for use with the "Alpaca (with RAW inputs)" chat template, though it also supports the Mistral V3 optional chat template.

Good for:

  • Roleplay (RP) Applications: Specifically engineered by merging models known for their "RP Uncensored" capabilities.
  • Creative Writing: The emphasis on RP core logic and substance core vocabulary suggests suitability for generating detailed and engaging narratives.
  • Extended Context Tasks: Its support for a 128K context size makes it suitable for applications requiring long-form memory and coherence.

Finetunes
2 models