BlogLLM Benchmark Scores as Published, Side by Side

Qwen3.5-27B: Benchmark Scores as Published

Every result the catalog holds for Qwen3.5-27B: 101 tests in 12 areas, from 4 pages. All were printed by Alibaba itself. Each number is as printed; nothing is ranked or rescored.

Developer AlibabaReleased 2026-02-24Context window 262KInput Text, image, videoOutput TextWeights Open (Apache 2.0)Data version 2026103001, as of 2026-10-30

Agentic Coding

TestValuePrinted bySetup as printed
SWE-bench Verified72.4%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
SWE-Bench Pro51.2%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more pageHarness: internal scaffold; Context: 200000
SWE-bench Multilingual69.3%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more pageHarness: internal scaffold; Context: 200000
Terminal-Bench 2.041.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
NL2Repo-Bench27.3Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more pageHarness: Claude Code (baselines)

Science and Math

TestValuePrinted bySetup as printed
GPQA Diamond85.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
HMMT February 202592%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
IMO-AnswerBench79.9%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page-
AIME 202692.6%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page-
HMMT November 202589.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
HMMT February 202684.3%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page-
PolyMATH71.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-

Long Context

TestValuePrinted bySetup as printed
AA-LCR66.1%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
LongBench v260.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-

Computer Use

TestValuePrinted bySetup as printed
OSWorld-Verified56.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
ScreenSpot-Pro70.3%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
AndroidWorld64.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-

Multimodal Understanding

TestValuePrinted bySetup as printed
CharXiv Reasoning (no tools)79.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MMMU-Pro75%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MMMU (validation)82.3%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MathVision86%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
MathVista (testmini)87.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
AI2D (test)92.9%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
BabyVision (with tools)44.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; code interpreter
LVBench73.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
OmniDocBench 1.588.9%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
Video-MMMU82.3%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
OCRBench89.4%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
HallusionBench70%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
MMStar81%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
BabyVision (no tools)34.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: none
ERQA60.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
SimpleVQA56%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
Video-MME (with subtitles)87%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
RealWorldQA83.7%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MMVU73.3%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
V* Bench (tool setting not stated)93.7%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15-
CC-OCR81%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MMBench-EN dev v1.192.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
CountBench97.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
DynaMath87.7%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
ZeroBench (subquestions)36.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
EmbSpatial-Bench84.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MLVU85.9%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
MMLongBench-Doc60.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
Video-MME (without subtitles)82.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
ZeroBench (metric not stated)10%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
MVBench74.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
RefSpatial-Bench67.7%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
LingoQA82%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
V* Bench (with tools)93.7%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; code interpreter
RefCOCO (average)90.9%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
VLMs are Blind96.9%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
ODinW-1341.1%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
Hypersim (metric not stated)13Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
nuScenes (metric not stated)15.2Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
SUN RGB-D (metric not stated)35.4Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
V* Bench (no tools)89%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: none
TIR-Bench (no tools)42.3%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: none
TIR-Bench (with code interpreter)59.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; code interpreter

General Knowledge and Reasoning

TestValuePrinted bySetup as printed
Humanity's Last Exam (no tools)24.3%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pagesTools: none
MMLU-Pro86.1%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
Humanity's Last Exam (with tools)48.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; search; Context: 256000
MMLU-Redux93.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
C-Eval90.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
SuperGPQA65.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-

Code Generation

TestValuePrinted bySetup as printed
LiveCodeBench v680.7%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 2 more pages-
Codeforces rating1899Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
OJBench (language not stated)40.1%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
QwenWebBench rating1068Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page-
FullStackBench en60.1%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
FullStackBench zh57.4%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-

Tool Use and Workflows

TestValuePrinted bySetup as printed
Toolathlon31.5%Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15-
MCP Atlas (public set)68.4%Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15-
MCPMark36.3%Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15-
τ²-bench (domain not stated)79%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
Claw-Eval pass^346.2%Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page-
DeepPlanning22.6%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
VitaBench41.9%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more page-
τ³-bench (domain not stated)68.4%Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15-
BFCL v468.5%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
SkillsBench27.2Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more pageHarness: OpenCode
Claw-Eval average score64.3Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page-
QwenClawBench52.2Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more pageContext: 256000

Web Search and Research

TestValuePrinted bySetup as printed
BrowseComp61%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; search; Context: 256000
BrowseComp-ZH62.1%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; search; Context: 256000
WideSearch (metric not stated)61.1Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15; printed on 1 more pageTools: enabled; search; Context: 256000
Seal-047.2%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15Tools: enabled; search; Context: 256000

Instruction Following

TestValuePrinted bySetup as printed
IFBench76.5Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
MultiChallenge60.8%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
IFEval95%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-

Multilingual

Healthcare

TestValuePrinted bySetup as printed
MedXpertQA MM62.4%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
PMC-VQA62.4%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-
SLAKE80%Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15-

Price

API list prices in US dollars per one million tokens, as printed on the developer's pricing page on the day checked.

PriceValuePrinted bySetup as printed
Input$0.086Alibaba, Alibaba Model Studio pricingchecked 2026-10-15Context: up to 128,000 tokens; Long context: $0.258 per 1M tokens above 128,000 tokens
Output$0.688Alibaba, Alibaba Model Studio pricingchecked 2026-10-15Context: up to 128,000 tokens; Long context: $2.064 per 1M tokens above 128,000 tokens

More from Alibaba

Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.