Skip to main content

Model profile

Qwen3.6-27B

AlibabaCurrentReleased Apr 21, 2026
Data verified
Overall Score
53.82Public #93 of 200
Arena Elo
Not listed
Eligible category ranks
4of 8
Price (1M tokens)
$0 in / $0 out
API pricing
Speed
Not listed
Context
262K

Evidence coverage

54 of 323 tracked benchmarks are published. 38 are verified and 16 provisional. 7 of 8 categories are measured.

Updated July 21, 2026Methodology
Published / tracked
54 / 323
Verified
38
Provisional
16
Categories with evidence
7 / 8

Evidence by category

  • Agentic10 benchmarks
    Mixed evidence
  • Coding8 benchmarks
    Mixed evidence
  • Reasoning2 benchmarks
    Reported
  • Knowledge12 benchmarks
    Mixed evidence
  • Math5 benchmarks
    Verified
  • Multilingual0 benchmarks
    Not measured
  • Multimodal16 benchmarks
    Mixed evidence
  • Inst. Following1 benchmark
    Reported
Open WeightSelf-hostReasoning
Confidence:
High
base

Qwen3.6-27B ranks #93 out of 200 models on the public leaderboard with an overall score of 53.82/100. It does not yet have enough sourced coverage for BenchLM's verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

Qwen3.6-27B is a open weight model with a 262K token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.

This profile currently has 54 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Multimodal & Grounded (#20), while its weakest is Agentic (#111). This performance profile makes it particularly strong for screenshots, documents, charts, and grounded multimodal workflows.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 53.4354.06

  1. MiMo-V2-Flash
    Xiaomi
    #9154.06
    MiMo-V2-Flash is #91 with a score of 54.06.
    Compare
  2. DeepSeek V4 Flash (High)
    DeepSeek
    #9253.95
    DeepSeek V4 Flash (High) is #92 with a score of 53.95.
    Compare
  3. Qwen3.6-27BCurrent model
    Alibaba
    #9353.82
    Qwen3.6-27B is #93 with a score of 53.82.
  4. GPT-5.1
    OpenAI
    #9453.65
    GPT-5.1 is #94 with a score of 53.65.
    Compare
  5. DeepSeek V3.1
    DeepSeek
    #9553.64
    DeepSeek V3.1 is #95 with a score of 53.64.
    Compare
  6. Claude Sonnet 4.5
    Anthropic
    #9653.61
    Claude Sonnet 4.5 is #96 with a score of 53.61.
    Compare
  7. DeepSeek V3.1 (Reasoning)
    DeepSeek
    #9753.43
    DeepSeek V3.1 (Reasoning) is #97 with a score of 53.43.
    Compare

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Multimodal32%
    Eligible cohort rank #20 of 29Category score 54.2
  2. Knowledge6%
    Eligible cohort rank #49 of 52Category score 50.8
  3. Coding28%
    Eligible cohort rank #88 of 122Category score 46.6
  4. Agentic7%
    Eligible cohort rank #111 of 119Category score 32.5

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #111 of 119Percentile 7thWeight 22%10 benchmarksMixed sources32.5
CodingRank #88 of 122Percentile 28thWeight 20%8 benchmarksMixed sources46.6
ReasoningWeight 17%2 benchmarksReportedScore pending
KnowledgeRank #49 of 52Percentile 6thWeight 12%12 benchmarksMixed sources50.8
MathRank Not rankedWeight 5%5 benchmarksVerified73.0
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalRank #20 of 29Percentile 32ndWeight 12%16 benchmarksMixed sources54.2
Inst. FollowingWeight 5%1 benchmarkReportedScore pending

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

Qwen3.6-27B
API / mo$0
Self-host / mo$429
Break-even
Model the full break-even

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic10 benchmarks
Terminal-Bench 2.0Provider exact
59.3%Weighted 38%
Source: Qwen3.6-27B model cardProvenance: Provider exact
Claw-EvalProvider exact
72.4%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
QwenClawBenchProvider exact
53.4%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
QwenWebBenchProvider exact
1487Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
AndroidWorldProvider exact
70.3%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
AA Agentic IndexReported

Artificial Analysis Agentic Index

27.0%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

94.2%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
GDPval-AAReported

GDPval-AA normalized

32.0%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
GDPval-AAReported
1140Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Gert LabsBenchmark exact

Gert Labs Composite Game Benchmark

54.84%Display only
Source: Gert Labs rankingsProvenance: Gert Labs reports this composite leaderboard score in the public rankings API. BenchLM scales the source gscore from 0-1 to 0-100 and stores it as a display-only agentic benchmark.
Coding8 benchmarks
LiveCodeBenchProvider exact

LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code

83.9%Weighted 38%
Source: Qwen3.6-27B model cardProvenance: Provider exact
SWE-bench VerifiedProvider exact

Software Engineering Benchmark Verified

77.2%Weighted 16%
Source: Qwen3.6-27B model cardProvenance: Provider exact
SWE-bench ProProvider exact
53.5%Weighted 10%
Source: Qwen3.6-27B model cardProvenance: Provider exact
SWE MultilingualProvider exact
71.3%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
Terminal-Bench 2.0Provider exact
59.3%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
NL2RepoProvider exact
36.2%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
AA Coding IndexReported

Artificial Analysis Coding Index

53.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-SciCodeReported

Artificial Analysis SciCode

39.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Reasoning2 benchmarks
AA-LCRReported

Artificial Analysis Long Context Reasoning

68.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
CritPtReported

Critical Physics Tasks

1.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Knowledge12 benchmarks
HLEProvider exact

Humanity's Last Exam

24%Weighted 45%
Source: Qwen3.6-27B model cardProvenance: Provider exact
MMLU-ProProvider exact

Massive Multitask Language Understanding Professional

86.2%Weighted 30%
Source: Qwen3.6-27B model cardProvenance: Provider exact
SuperGPQAProvider exact

SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines

66%Weighted 7%
Source: Qwen3.6-27B model cardProvenance: Provider exact
GPQAProvider exact

Graduate-Level Google-Proof Q&A

87.8%Weighted 7%
Source: Qwen3.6-27B model cardProvenance: Provider exact
MMLU-ReduxProvider exact
93.5%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
C-EvalProvider exact
91.4%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
Artificial Analysis Intelligence IndexReported
37.0%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-GPQA DiamondReported

Artificial Analysis GPQA Diamond

84.2%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-HLEReported

Artificial Analysis Humanity's Last Exam

21.6%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience IndexReported

Artificial Analysis Omniscience Index

-19.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience AccuracyReported

Artificial Analysis Omniscience Accuracy

19.2%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience Hallucination RateReported

Artificial Analysis Omniscience Hallucination Rate

48.3%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Math5 benchmarks
HMMT Feb 2026Provider exact

Harvard-MIT Mathematics Tournament February 2026

84.3%Weighted 25%
Source: Qwen3.6-27B model cardProvenance: Provider exact
AIME26Provider exact

AIME 2026

94.1%Weighted 25%
Source: Qwen3.6-27B model cardProvenance: Provider exact
HMMT Feb 2025Provider exact

Harvard-MIT Mathematics Tournament February 2025

93.8%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
HMMT Nov 2025Provider exact

Harvard-MIT Mathematics Tournament November 2025

90.7%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
MMAnswerBenchProvider exact
80.8%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
Multimodal16 benchmarks
MMMU-ProProvider exact

Massive Multi-discipline Multimodal Understanding Pro

75.8%Weighted 45%
Source: Qwen3.6-27B model cardProvenance: Provider exact
CharXivProvider exact

CharXiv Reasoning

78.4%Weighted 25%
Source: Qwen3.6-27B model cardProvenance: Provider exact
MMMUProvider exact

Massive Multi-discipline Multimodal Understanding

82.9%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
RealWorldQAProvider exact
84.1%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
DynaMathProvider exact
85.6%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
MStarProvider exact
81.4%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
SimpleVQAProvider exact
56.1%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
CC-OCRProvider exact
81.2%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
CountBenchProvider exact
97.8%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
RefCOCO (avg)Provider exact

RefCOCO average

92.5%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
ERQAProvider exact
62.5%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
Video-MME (with subtitle)Provider exact

Video-MME with subtitle

87.7%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
VideoMMMUProvider exact
84.4%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
MLVU (M-Avg)Provider exact

MLVU mean average

86.6%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
V*Provider exact
94.7%Display only
Source: Qwen3.6-27B model cardProvenance: Provider exact
AA-MMMU-ProReported

Artificial Analysis MMMU-Pro

74.6%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Inst. Following1 benchmark
AA-IFBenchReported

Artificial Analysis IFBench

67.6%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.

Frequently Asked Questions

How does Qwen3.6-27B perform overall in AI benchmarks?

Qwen3.6-27B currently ranks #93 out of 200 models on BenchLM's provisional leaderboard with an overall score of 53.82. It is created by Alibaba. Its published context window is 262K.

Is Qwen3.6-27B good for knowledge and understanding?

Qwen3.6-27B ranks #49 out of 52 models in knowledge and understanding benchmarks with an average score of 50.8. There are stronger options in this category.

Is Qwen3.6-27B good for coding and programming?

Qwen3.6-27B ranks #88 out of 122 models in coding and programming benchmarks with an average score of 46.6. There are stronger options in this category.

Is Qwen3.6-27B good for mathematics?

Qwen3.6-27B has visible benchmark coverage in mathematics, but BenchLM does not currently assign it a global category rank there.

Is Qwen3.6-27B good for reasoning and logic?

Qwen3.6-27B has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.

Is Qwen3.6-27B good for agentic tool use and computer tasks?

Qwen3.6-27B ranks #111 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 32.5. There are stronger options in this category.

Is Qwen3.6-27B good for multimodal and grounded tasks?

Qwen3.6-27B ranks #20 out of 29 models in multimodal and grounded tasks benchmarks with an average score of 54.2. There are stronger options in this category.

Is Qwen3.6-27B good for instruction following?

Qwen3.6-27B has visible benchmark coverage in instruction following, but BenchLM does not currently assign it a global category rank there.

Is Qwen3.6-27B open source?

Yes, Qwen3.6-27B is an open weight model created by Alibaba, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

Does Qwen3.6-27B have full benchmark coverage on BenchLM?

Not yet. Qwen3.6-27B currently has 54 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of Qwen3.6-27B?

Qwen3.6-27B has a published context window of 262K, which determines how much text it can process in a single interaction.

Last updated: July 21, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.