What are the hardware requirements for Ninja V1 RP?

Ninja V1 RP requires approximately 14.4 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Ninja V1 RP and how large is it?

Ninja V1 RP is developed by Aratako, a model with 7.2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Ninja V1 RP?

Ninja V1 RP is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Ninja V1 RP by Aratako — VRAM 14.4GB, 4K context

Name: Ninja V1 RP
Author: Aratako

Ninja V1 RP is an open-source language model by Aratako. Features: 7.2b LLM, VRAM: 14.4GB, Context: 4K, License: cc-by-nc-4.0, Merged, LLM Explorer Score: 0.12.

Merged Model Base model:aratako/ninja-v1-rp... Base model:finetune:aratako/ni... Dataset:aratako/bluemoon top50... Dataset:aratako/limarp-augment... Dataset:aratako/rosebleu-1on1-... Dataset:grimulkan/limarp-augme... Dataset:omniaicreator/japanese... Dataset:sicariussicariistuff/b... Endpoints compatible Ja Mistral Region:us Roleplay Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Aratako/Ninja-v1-RP

Ninja V1 RP Parameters and Internals

Model Type

roleplay

Supported Languages

ja (Full)

Training Details

Data Sources:

Aratako/Rosebleu-1on1-Dialogues-RP, Aratako/LimaRP-augmented-ja-karakuri, grimulkan/LimaRP-augmented, Aratako/Bluemoon_Top50MB_Sorted_Fixed_ja, SicariusSicariiStuff/Bluemoon_Top50MB_Sorted_Fixed, OmniAICreator/Japanese-Roleplay

Methodology:

Merge of pre-trained language models using mergekit. The task vectors from four models were multiplied by 0.8 and added to the base model.

Input Output

Input Format:

Vicuna chat template. System prompts should be before the first `USER:` input.

Performance Tips:

Each turn should end with an `eos_token` (`~~`).

LLM Name	Ninja V1 RP
Repository 🤗	https://huggingface.co/Aratako/Ninja-v1-RP
Base Model(s)	Ninja V1 RP WIP Aratako/Ninja-v1-RP-WIP
Merged Model	Yes
Model Size	7.2b
Required VRAM	14.4 GB
Updated	2026-08-24
Maintainer	Aratako
Model Type	mistral
Model Files	9.9 GB: 1-of-2 4.5 GB: 2-of-2
Supported Languages	ja
Model Architecture	MistralForCausalLM
License	cc-by-nc-4.0
Context Length	4096
Model Max Length	4096
Transformers Version	4.40.0.dev0
Tokenizer Class	LlamaTokenizer
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the Ninja V1 RP capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Ninja V1 RP

Best Alternatives	Context / RAM	Downloads	Likes
...Web AI HumanAI 012 INSTRUCT XA	512K / 14.4 GB	9	0
...Web AI HumanAI 012 INSTRUCT IA	512K / 14.4 GB	8	0
...Web AI HumanAI 011 INSTRUCT ML	512K / 14.4 GB	7	0
14	512K / 14.5 GB	319	0
SpydazWeb HumanAI M2	512K / GB	38	0
...dazWeb AI HumanAI 011 INSTRUCT	512K / 14.4 GB	14	1
SpydazWeb AI HumanAI 009 CHAT	512K / 14.4 GB	6	0
SpydazWeb AI HumanAI 007	512K / 14.4 GB	5	0
SpydazWebAI Human AGI 001	512K / 14.4 GB	6	1
...Web AI HumanAI 012 INSTRUCT MX	512K / 14.4 GB	8	0

Note: green Score (e.g. "73.2") means that the model is better than Aratako/Ninja-v1-RP.