What are the hardware requirements for Saka 24B?

Saka 24B requires approximately 47.3 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Saka 24B and how large is it?

Saka 24B is developed by Sakalti, a model with 24b 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 Saka 24B?

Saka 24B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Saka 24B by Sakalti — VRAM 47.3GB, 32K context

Name: Saka 24B
Author: Sakalti

Saka 24B is an open-source language model by Sakalti. Features: 24b LLM, VRAM: 47.3GB, Context: 32K, Instruction-Based, Merged, LLM Explorer Score: 0.25.

Merged Model Arxiv:2408.07990 Autotrain compatible Base model:alsamcur123/mistral... Base model:huihui-ai/mistral-s... Base model:trashpanda-org/ms-2... Conversational Endpoints compatible Instruct Mistral Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/Sakalti/Saka-24B

Saka 24B Parameters and Internals

LLM Name	Saka 24B
Repository 🤗	https://huggingface.co/Sakalti/Saka-24B
Base Model(s)	... 24B Instruct 2501 Abliterated MS 24B Instruct Mullein V0 ...all3 24B InstructContinuedFine mergekit-community/24B-MS-PRO-V0.01 huihui-ai/Mistral-Small-24B-Instruct-2501-abliterated trashpanda-org/MS-24B-Instruct-Mullein-v0 AlSamCur123/Mistral-Small3-24B-InstructContinuedFine mergekit-community/24B-MS-PRO-V0.01
Merged Model	Yes
Model Size	24b
Required VRAM	47.3 GB
Updated	2025-02-12
Maintainer	Sakalti
Model Type	mistral
Instruction-Based	Yes
Model Files	4.9 GB: 1-of-10 4.8 GB: 2-of-10 4.8 GB: 3-of-10 4.9 GB: 4-of-10 4.8 GB: 5-of-10 4.8 GB: 6-of-10 4.9 GB: 7-of-10 4.8 GB: 8-of-10 4.8 GB: 9-of-10 3.8 GB: 10-of-10
Model Architecture	MistralForCausalLM
Context Length	32768
Model Max Length	32768
Transformers Version	4.48.2
Tokenizer Class	LlamaTokenizer
Vocabulary Size	131072
Torch Data Type	bfloat16

Rank the Saka 24B 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 Saka 24B

Best Alternatives	Context / RAM	Downloads	Likes
...ral Nemo Instruct 24B Merge V1	1000K / 24.6 GB	9	0
...all 3.1 24B Instruct 2503 Text	128K / 47.3 GB	16	0
Cydonia 24B V4.3	128K / 47.3 GB	3652	138
Slimaki 24B V1.2	128K / 47.3 GB	28	6
Neural V1 24B	128K / 47.3 GB	9	3
MS3.2 PaintedFantasy V4.1 24B	128K / 47.3 GB	80	18
Cydonia 24B V4.1	128K / 47.3 GB	2162	80
Slimaki 24B V1	128K / 47.3 GB	15	10
Asmodeus 24B V1	128K / 47.3 GB	50	13
MS3.2 PaintedFantasy V3 24B	128K / 47.3 GB	28	35

Note: green Score (e.g. "73.2") means that the model is better than Sakalti/Saka-24B.