What are the hardware requirements for K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8?

K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 requires approximately 44.5 GB of VRAM and supports a context window of 512K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 and how large is it?

K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 is developed by primitive-ai, a model with 36b 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 K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8?

K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 by primitive-ai — VRAM 44.5GB, 512K context

Name: K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8
Author: primitive-ai

K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 is an open-source language model by primitive-ai. Features: 36b LLM, VRAM: 44.5GB, Context: 512K, License: apache-2.0, LLM Explorer Score: 0.26.

Base model:ifm/k2-horizon-mova... Base model:quantized:ifm/k2-ho... Blackwell Compressed-tensors Conversational Custom code Fp8 K2-horizon K2 horizon Mixed-precision Moe Mova Nvfp4 Quantized Reasoning Region:us Safetensors Sharded Tensorflow Tool-calling Vllm

Model Card on HF 🤗: https://huggingface.co/primitive-ai/K2-Horizon-MoVA-36B-A4B-mixed-NVFP4-FP8

K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 Parameters and Internals

LLM Name	K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8
Repository 🤗	https://huggingface.co/primitive-ai/K2-Horizon-MoVA-36B-A4B-mixed-NVFP4-FP8
Base Model(s)	K2 Horizon MoVA 36B A4B IFM/K2-Horizon-MoVA-36B-A4B
Model Size	36b
Required VRAM	44.5 GB
Updated	2026-09-08
Maintainer	primitive-ai
Model Type	k2_horizon
Model Files	5.0 GB: 1-of-9 5.0 GB: 2-of-9 5.0 GB: 3-of-9 5.0 GB: 4-of-9 5.0 GB: 5-of-9 5.0 GB: 6-of-9 5.0 GB: 7-of-9 5.0 GB: 8-of-9 4.5 GB: 9-of-9
Model Architecture	K2HorizonForCausalLM
License	apache-2.0
Context Length	524288
Model Max Length	524288
Transformers Version	5.15.0
Tokenizer Class	TokenizersBackend
Vocabulary Size	250624

Rank the K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8 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 K2 Horizon MoVA 36B A4B Mixed NVFP4 FP8

Best Alternatives	Context / RAM	Downloads	Likes
K2 Horizon MoVA 36B A4B	512K / 63 GB	433	139
Jais 3 MoVA 36B A4B	512K / 149.8 GB	1569	9
K2 Horizon MoVA 36B A4B MLA	512K / 64.6 GB	536	1
...zon MoVA 36B A4B DERISKED BF16	512K / 64.6 GB	524	3
K2 Horizon MoVA 36B A4B OQ4e	512K / 22 GB	1098	2
...on MoVA 36B A4B DERISKED NVFP4	512K / 33.8 GB	242	1
...2 Horizon MoVA 36B A4B Heretic	512K / 74.9 GB	210	1
K2 Horizon MoVA 36B A4B FP8	512K / 40.6 GB	129	6
K2 Horizon MoVA 36B A4B 4bit	512K / 26.5 GB	315	3
... Horizon MoVA 36B A4B MLX 4bit	512K / 15.6 GB	692	3

Note: green Score (e.g. "73.2") means that the model is better than primitive-ai/K2-Horizon-MoVA-36B-A4B-mixed-NVFP4-FP8.