What are the hardware requirements for Sage A1.5 FP8?

Sage A1.5 FP8 requires approximately 37.7 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Sage A1.5 FP8 and how large is it?

Sage A1.5 FP8 is developed by InternScience, 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 Sage A1.5 FP8?

Sage A1.5 FP8 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Sage A1.5 FP8 by InternScience — VRAM 37.7GB

Name: Sage A1.5 FP8
Author: InternScience

Sage A1.5 FP8 is an open-source language model by InternScience. Features: 36b LLM, VRAM: 37.7GB, License: apache-2.0, LLM Explorer Score: 0.36.

Arxiv:2610.28578 Agentic Base model:internscience/sage-... Base model:quantized:internsci... Code Compressed-tensors Conversational Endpoints compatible Image-text-to-text Moe Qwen3 5 moe Region:us Safetensors Sharded Tensorflow Tool-use Vision Vlm

Model Card on HF 🤗: https://huggingface.co/InternScience/Sage-A1.5-FP8

Sage A1.5 FP8 Parameters and Internals

LLM Name	Sage A1.5 FP8
Repository 🤗	https://huggingface.co/InternScience/Sage-A1.5-FP8
Base Model(s)	InternScience/Sage-A1.5 InternScience/Sage-A1.5
Model Size	36b
Required VRAM	37.7 GB
Updated	2026-10-27
Maintainer	InternScience
Model Type	qwen3_5_moe
Model Files	20.0 GB: 1-of-2 17.7 GB: 2-of-2 1.7 GB
Model Architecture	Qwen3_5MoeForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	4.57.1
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Errors	replace

Rank the Sage A1.5 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 Sage A1.5 FP8

Best Alternatives	Context / RAM	Downloads	Likes
Sage A1.5	0K / 70.6 GB	2033	266
...c Quality Inspector V1.0 NVFP4	0K / 25.6 GB	316	0
Apodex 1.1 Mini FP8	0K / 37.6 GB	235	14
Occamy 1.0 OQ6e Mtp	0K / 30.2 GB	43	1
Affine 5eqdtdzqle Ckpt300 M4	0K / 142.6 GB	548	0
AdvancedMathBench AutoVerifier	0K / 0.1 GB	50	1
Darwin 35B A3B Opus	0K / 72.3 GB	194	97
Yaya 3.6	0K / 72.2 GB	38	0
Carnice Qwen3.6 MoE 35B A3B	0K / 72.2 GB	0	23
Qwopus MoE 35B A3B	0K / 72.3 GB	3	11

Note: green Score (e.g. "73.2") means that the model is better than InternScience/Sage-A1.5-FP8.