- Context
- 262Ktokens
- Released
- 8mo ago
- Size
- 4Bparameters
Qwen3.5 4B is a model in the Qwen3.5 series, which Alibaba announced as its latest model series with an open-source release. Alibaba's launch announcement presents Qwen3.5 as a natively multimodal model built for high-efficiency inference.
- Multimodal model
- Open weights · 4B
- Accepts images, video
- Reasoning
- Tools
How it compares
Size class
Tiny4BAn open-weight model under 5 billion parameters, small enough for phones, laptops and edge devices.
Open models of a similar size from other labs
Vs. your stack
Add the models you use to see how this one compares.
Sign inAdditional source evidence
- Source model
- alibaba/qwen3.5-4b
- Source family
- qwen
- Last updated
- 2026-03-02
- Model context
- 262K
- Model output limit
- 82K
- Open weights
- Yes
- File attachments
- Yes
- Reasoning
- Yes
- Tool calling
- Yes
- Temperature control
- Yes
Downloadable weights
Complete source record (JSON)
{
"id": "alibaba/qwen3.5-4b",
"name": "Qwen3.5 4B",
"description": "Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use",
"family": "qwen",
"last_updated": "2026-03-02",
"attachment": true,
"reasoning": true,
"tool_call": true,
"temperature": true,
"modalities": {
"input": [
"text",
"image",
"video"
],
"output": [
"text"
]
},
"limit": {
"context": 262144,
"output": 81920
},
"release_date": "2026-03-02",
"open_weights": true,
"license": "Apache-2.0",
"weights": [
{
"label": "Hugging Face",
"url": "https://huggingface.co/Qwen/Qwen3.5-4B"
}
]
}Frequently asked questions
Running locally
At Q4 it needs an estimated 3.3 GB of VRAM, at Q8 an estimated 5.4 GB and at FP16 an estimated 9.7 GB. All figures are estimates from the 4B parameter count, covering weights plus about 15% and 0.5 GB for a short context.
Yes. A 24 GB, 16 GB or 12 GB GPU fits it at FP16 (estimated 9.7 GB), and an 8 GB GPU fits it at Q8 (estimated 5.4 GB). All figures are estimates.
Yes. At Q4 it needs an estimated 3.3 GB, and assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits Q4 is 8 GB (estimate).
Privacy and data
Yes. Qwen3.5 4B has open weights, so you can self-host it for full control of your data.
Capabilities and specs
The context window is 262K tokens (262,144) and the maximum output is 82K tokens (81,920).
Yes to images and video, no to audio. Its input modalities are text, image and video, and PDFs are not listed as a supported input.
Yes, tool calling (function calling) is supported.
Yes, a reasoning mode is supported.
Quality and benchmarks
No. Compared with its predecessor Qwen3 8B, Qwen3.5 4B is ahead on none, behind on OTIS Mock AIME 2024-2025 and level on none.
Matched on model kind and release date, with no comparable price: Olmo Hybrid 7B Instruct by Ai2, Olmo Hybrid 7B Think by Ai2 and Nemotron Cascade 2 30B A3B by NVIDIA.
Olmo Hybrid 7B InstructQwen3.5 4B vs Olmo Hybrid 7B InstructOlmo Hybrid 7B ThinkQwen3.5 4B vs Olmo Hybrid 7B ThinkNemotron Cascade 2 30B A3BQwen3.5 4B vs Nemotron Cascade 2 30B A3B
Versions and lineage
Their listed context windows and release dates are the same. Both have a 262K token (262,144) context window and both were released on March 2, 2026.
Licensing and access
Qwen3.5 4B has open weights, released under the Apache-2.0 license.
Yes. The license is Apache 2.0, a permissive license that allows commercial use.
Basics
Qwen made Qwen3.5 4B, and it was released on March 2, 2026.
Qwen3.5 4B has open weights.
Generated from Vertical Tabs data, updated November 10, 2026.
Discussion
No comments yet
Be the first to share your thoughts.