What are the hardware requirements for Mellum2 12B A2.5B Instruct?

Mellum2 12B A2.5B Instruct requires approximately 24.3 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Mellum2 12B A2.5B Instruct and how large is it?

Mellum2 12B A2.5B Instruct is developed by JetBrains, a model with 12b 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 Mellum2 12B A2.5B Instruct?

Mellum2 12B A2.5B Instruct is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Mellum2 12B A2.5B Instruct by JetBrains — VRAM 24.3GB, 128K context

Name: Mellum2 12B A2.5B Instruct
Author: JetBrains

Mellum2 12B A2.5B Instruct is an open-source language model by JetBrains. Features: 12b LLM, VRAM: 24.3GB, Context: 128K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.32.

Arxiv:2605.31268 Conversational En Endpoints compatible Eval-results Instruct Mellum Model-index Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/JetBrains/Mellum2-12B-A2.5B-Instruct

Mellum2 12B A2.5B Instruct Parameters and Internals

LLM Name	Mellum2 12B A2.5B Instruct
Repository 🤗	https://huggingface.co/JetBrains/Mellum2-12B-A2.5B-Instruct
Model Size	12b
Required VRAM	24.3 GB
Updated	2026-07-25
Maintainer	JetBrains
Model Type	mellum
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-5 5.0 GB: 2-of-5 5.0 GB: 3-of-5 5.0 GB: 4-of-5 4.3 GB: 5-of-5
Supported Languages	en
Model Architecture	MellumForCausalLM
License	apache-2.0
Context Length	131072
Model Max Length	131072
Transformers Version	5.8.1
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|endoftext\|>
Vocabulary Size	98304

Rank the Mellum2 12B A2.5B Instruct 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

Quantized Models of the Mellum2 12B A2.5B Instruct

Model	Likes	Downloads	VRAM
...ellum2 12B A2.5B Instruct 4bit	1	14	6 GB

Best Alternatives to Mellum2 12B A2.5B Instruct

Best Alternatives	Context / RAM	Downloads	Likes
...llum2 12B A2.5B Instruct Mxfp4	128K / 6.4 GB	166	1
...ellum2 12B A2.5B Instruct 4bit	128K / 6.8 GB	14	1