Qwen3.5 4B is a compact open-weight multimodal model from Alibaba for reasoning, coding, visual understanding, tool use, and structured output.
Added Aug 16, 2026
Context Window
262.1K
Max Output
32.8K
Input Price (Auto)
$0.10/1M
Output Price (Auto)
$0.20/1M
Cache Read (Auto)
$0.050/1M
Capabilities
Benchmarks
Benchmarks
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
16.1
Coding Index
20.3
Reasoning
GPQA Diamond
Graduate-level scientific reasoning
71.2%
Better than 52% of models compared
HLE
Humanity's Last Exam
8.0%
Better than 52% of models compared
IFBench
Instruction-following benchmark
33.3%
Better than 22% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
87.7%
Better than 82% of models compared
AA-LCR
Long context reasoning evaluation
34.0%
Better than 40% of models compared
Coding
SciCode
Python programming for scientific computing
18.3%
Better than 18% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
11.4%
Better than 47% of models compared
Last updated Aug 18, 2026
Artificial AnalysisProviders
Choose explicit providers for this model. Auto routing remains available as the default option.
Loading provider options…
Related text models
Compare Qwen3.5 4B with similar models from the same provider or model family.
Qwen 3.8 27B Fable
qwen/qwen3.8-27b-fableQwen 3.8 27B Fable is an open-weight multimodal creative finetune for expressive dialogue, long-form storytelling, character work, and roleplay.
Qwen3.8 Flash
alibaba/qwen3.8-flashQwen3.8 Flash is Alibaba's latest fast multimodal model, with a million-token context window for coding, agentic workflows, visual understanding, long documents, codebases, and videos.
Qwen3.8 2.4T A95B (Max)
qwen/qwen3.8-2.4t-a95bThis is the same underlying model as Qwen3.8 Max, exposed under its architecture-based 2.4T A95B name for easier discovery. It uses the identical routing, pricing, capabilities, and non-thinking mode.
Qwen 3.8 27B Uncensored Thinking
qwen/qwen3.8-27b-uncensored:thinkingQwen 3.8 27B Uncensored with thinking enabled for more deliberate creative work, coding, multimodal analysis, tool use, and long-context problem solving.
Qwen 3.6 35B A3B Uncensored Thinking
qwen/qwen3.6-35b-a3b-uncensored:thinkingQwen 3.6 35B A3B Uncensored with thinking enabled for more deliberate coding, multimodal analysis, tool use, and complex chat tasks.
Qwen 3.8 27B Obliterated
qwen/qwen3.8-27b-obliteratedQwen 3.8 27B Obliterated is an open-weight multimodal model LoRA-tuned for fewer refusals across chat, coding, reasoning, tool use, and long-context work.