What are the hardware requirements for Step3 VL 10B Base?

Step3 VL 10B Base requires approximately 3.2 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Step3 VL 10B Base and how large is it?

Step3 VL 10B Base is developed by stepfun-ai, a model with 10b 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 Step3 VL 10B Base?

Step3 VL 10B Base is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Step3 VL 10B Base by stepfun-ai — VRAM 3.2GB

Name: Step3 VL 10B Base
Author: stepfun-ai

Step3 VL 10B Base is an open-source language model by stepfun-ai. Features: 10b LLM, VRAM: 3.2GB, License: apache-2.0, LLM Explorer Score: 0.25.

Arxiv:2601.09668 Custom code Image-text-to-text Region:us Safetensors Step robotics

Model Card on HF 🤗: https://huggingface.co/stepfun-ai/Step3-VL-10B-Base

Step3 VL 10B Base Parameters and Internals

LLM Name	Step3 VL 10B Base
Repository 🤗	https://huggingface.co/stepfun-ai/Step3-VL-10B-Base
Model Size	10b
Required VRAM	3.2 GB
Updated	2026-08-08
Maintainer	stepfun-ai
Model Type	step_robotics
Model Files	4.3 GB 4.3 GB 4.2 GB 4.3 GB 3.2 GB
Model Architecture	StepVLForConditionalGeneration
License	apache-2.0
Model Max Length	131072
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Errors	replace

Rank the Step3 VL 10B Base 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 Step3 VL 10B Base

Best Alternatives	Context / RAM	Downloads	Likes
Step3 VL 10B	0K / 3.2 GB	138547	411

Note: green Score (e.g. "73.2") means that the model is better than stepfun-ai/Step3-VL-10B-Base.