Qwen3.5 MediumOpen Weight

Released

Qwen3.5 Medium is an AI model released by Qwen on Tuesday, Feb 24 2026, 8 days after Qwen3.5-Plus. It is an open-weight model — the trained weights are available to download and run. It was released in 27B and 35B and 122B parameter sizesThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality., with a 256k token context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider.. Benchmark results (shown below) cover IFEval, SWE-Bench Verified, LiveCodeBench, Terminal-Bench 2.0, BrowseComp, Humanity's Last Exam, and 7 more.

Benchmarks

Coding

SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better.
72%
#43 of 95
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better.
78.9%
#25 of 86

Terminal & CLI

Terminal-Bench 2.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? (Version 2.0 of the test.) Higher is better.
49.4%
#25 of 37

Agentic & tool use

IFEvalAgentic tasks — following verifiable instructions in generated responses
93.4%
#1
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better.
63.8%
#30 of 48
OSWorld-VerifiedAgentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better.
58%
#29 of 36

Reasoning & science

Humanity's Last ExamMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better.
25.3%no tools
47.5%with tools
no tools
#41 of 58
with tools
#40 of 55
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better.
86.6%
#45 of 145
IFBenchInstruction following — Tests whether the AI can follow detailed instructions and satisfy multiple constraints at once. Higher is better.
76.1%
#9 of 16

Knowledge work

MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark
86.7%
#4 of 58

Multimodal

CharXiv ReasoningChart reasoning — Can the AI read and reason about complex charts and figures, not just text? Higher is better.
77.2%
#24 of 32
MMMU-ProMultimodal reasoning — A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better.
76.9%
#18 of 36
MMMUMultimodal — Tests the AI on understanding images and text together across many college subjects. Higher is better.
83.9%
#4 of 50

Compare Qwen3.5 Medium with

Qwen3.5 Medium

About

The open-weight Qwen3.5 Medium release grouped the 27B dense and 35B-A3B and 122B-A10B mixture-of-experts checkpoints. It brought the February 2026 flagship's text and vision functions to three smaller deployment sizes.

The benchmark scores on this page come from the 122B-A10B checkpoint, which scored 86.6 on GPQA Diamond and 72.0 on SWE-bench Verified. Qwen documented Qwen3.5-Flash as a hosted counterpart to the 35B-A3B checkpoint, with a one-million-token default context and built-in tools.

Frequently asked questions

Qwen3.5 Medium was released by Qwen on Tuesday, Feb 24 2026.

All Qwen releases

42 tracked