What are the hardware requirements for Step 3.5 Flash FP8?

Step 3.5 Flash FP8 requires approximately 4.5 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Step 3.5 Flash FP8 and how large is it?

Step 3.5 Flash FP8 is developed by stepfun-ai, a model with 199.4b 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 Step 3.5 Flash FP8?

Step 3.5 Flash FP8 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Step 3.5 Flash FP8 by stepfun-ai — VRAM 4.5GB, 256K context

Name: Step 3.5 Flash FP8
Author: stepfun-ai

Step 3.5 Flash FP8 is an open-source language model by stepfun-ai. Features: 199.4b LLM, VRAM: 4.5GB, Context: 256K, License: apache-2.0, LLM Explorer Score: 0.25.

Arxiv:2507.19427 Arxiv:2601.05593 Arxiv:2602.10604 Conversational Custom code Fp8 Region:us Safetensors Step3p5

Model Card on HF 🤗: https://huggingface.co/stepfun-ai/Step-3.5-Flash-FP8

Step 3.5 Flash FP8 Parameters and Internals

LLM Name	Step 3.5 Flash FP8
Repository 🤗	https://huggingface.co/stepfun-ai/Step-3.5-Flash-FP8
Base Model(s)	stepfun-ai/step-3.5-flash stepfun-ai/step-3.5-flash
Model Size	199.4b
Required VRAM	4.5 GB
Updated	2026-08-08
Maintainer	stepfun-ai
Model Type	step3p5
Model Files	9.6 GB 8.6 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB 4.5 GB
Model Architecture	Step3p5ForCausalLM
License	apache-2.0
Context Length	262144
Model Max Length	262144
Vocabulary Size	128896
Torch Data Type	bfloat16

Rank the Step 3.5 Flash 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 Step 3.5 Flash FP8

Best Alternatives	Context / RAM	Downloads	Likes
Step 3.5 Flash	256K / 8.6 GB	167827	825

Note: green Score (e.g. "73.2") means that the model is better than stepfun-ai/Step-3.5-Flash-FP8.