Skip to main content

Model profile

Qwen3.5-35B-A3B

AlibabaCurrentReleased Mar 4, 2026
Data verified
Overall Score
56.97Public #72 of 200Verified #46 of 99
Arena Elo
1396
Eligible category ranks
6of 8
Price (1M tokens)
$0 in / $0 out
API pricing
Speed
Not listed
Context
262K

Evidence coverage

28 of 323 tracked benchmarks are published. 14 are verified and 14 provisional. 7 of 8 categories are measured.

Updated July 23, 2026Methodology
Published / tracked
28 / 323
Verified
14
Provisional
14
Categories with evidence
7 / 8

Evidence by category

  • Agentic5 benchmarks
    Mixed evidence
  • Coding3 benchmarks
    Mixed evidence
  • Reasoning3 benchmarks
    Mixed evidence
  • Knowledge9 benchmarks
    Mixed evidence
  • Math0 benchmarks
    Not measured
  • Multilingual1 benchmark
    Verified
  • Multimodal5 benchmarks
    Mixed evidence
  • Inst. Following2 benchmarks
    Mixed evidence
Open WeightSelf-hostReasoning
Confidence:
High
base

Qwen3.5-35B-A3B ranks #72 out of 200 models on the public leaderboard with an overall score of 56.97/100. It also ranks #46 out of 99 on the verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

Qwen3.5-35B-A3B 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 28 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 Multilingual (#11), while its weakest is Agentic (#74). This performance profile makes it a well-rounded choice across a range of tasks.

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 56.5957.25

  1. Gemini 2.5 Pro
    Google
    #7057.25
    Gemini 2.5 Pro is #70 with a score of 57.25.
    Compare
  2. Qwen3.5 397B
    Alibaba
    #7157.01
    Qwen3.5 397B is #71 with a score of 57.01.
    Compare
  3. Qwen3.5-35B-A3BCurrent model
    Alibaba
    #7256.97
    Qwen3.5-35B-A3B is #72 with a score of 56.97.
  4. GPT-5.3-Codex-Spark
    OpenAI
    #7356.91
    GPT-5.3-Codex-Spark is #73 with a score of 56.91.
    Compare
  5. Kimi K2.6
    Moonshot AI
    #7456.79
    Kimi K2.6 is #74 with a score of 56.79.
    Compare
  6. GPT-5.4 mini
    OpenAI
    #7556.77
    GPT-5.4 mini is #75 with a score of 56.77.
    Compare
  7. Grok 4 Fast (Reasoning)
    xAI
    #7656.59
    Grok 4 Fast (Reasoning) is #76 with a score of 56.59.
    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. Multilingual17%
    Eligible cohort rank #11 of 13Category score 21.1
  2. Inst. Following50%
    Eligible cohort rank #16 of 31Category score 84.5
  3. Knowledge67%
    Eligible cohort rank #18 of 52Category score 75.3
  4. Multimodal0%
    Eligible cohort rank #29 of 29Category score 0.0
  5. Coding49%
    Eligible cohort rank #63 of 122Category score 50.1
  6. Agentic38%
    Eligible cohort rank #74 of 119Category score 45.3

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 #74 of 119Percentile 38thWeight 22%5 benchmarksMixed sources45.3
CodingRank #63 of 122Percentile 49thWeight 20%3 benchmarksMixed sources50.1
ReasoningRank Not rankedWeight 17%3 benchmarksMixed sources72.0
KnowledgeRank #18 of 52Percentile 67thWeight 12%9 benchmarksMixed sources75.3
MathWeight 5%0 benchmarksNot measuredNot measured
MultilingualRank #11 of 13Percentile 17thWeight 7%1 benchmarkVerified21.1
MultimodalRank #29 of 29Percentile 0thWeight 12%5 benchmarksMixed sources0.0
Inst. FollowingRank #16 of 31Percentile 50thWeight 5%2 benchmarksMixed sources84.5

Chatbot Arena performance

Scroll horizontally to inspect confidence intervals and vote counts.

Chatbot Arena Elo, confidence interval, and vote count by evaluation view
ViewEloConfidence intervalVotes
Text Overall1396±4.329,157
Coding1435±7.37,988
Math1399±14.11,758
Instruction Following1388±6.79,282
Creative Writing1344±9.34,456
Multi-turn1395±8.75,204
Hard Prompts1413±5.318,332
Hard Prompts (English)1423±6.98,988
Longer Query1402±6.511,063

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.

Agentic5 benchmarks
Terminal-Bench 2.0Provider exact
40.5%Weighted 38%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
OSWorld-VerifiedProvider exact
54.5%Weighted 34%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
BrowseCompProvider exact
61%Weighted 28%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

89.2%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.
Gert LabsBenchmark exact

Gert Labs Composite Game Benchmark

28.96%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.
Coding3 benchmarks
SWE-RebenchBenchmark exact
53.7%Weighted 20%
Source: SWE-Rebench leaderboardProvenance: Live default SWE-Rebench leaderboard lists Qwen3.5-35B-A3B at 53.7% resolved rate.
SWE-bench VerifiedProvider exact

Software Engineering Benchmark Verified

69.2%Weighted 16%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
AA-SciCodeReported

Artificial Analysis SciCode

37.7%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.
Reasoning3 benchmarks
LongBench v2Provider exact
59%Weighted 38%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
AA-LCRReported

Artificial Analysis Long Context Reasoning

62.7%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

0.9%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.
Knowledge9 benchmarks
MMLU-ProProvider exact

Massive Multitask Language Understanding Professional

85.3%Weighted 30%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
SuperGPQAProvider exact

SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines

63.4%Weighted 7%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
GPQAProvider exact

Graduate-Level Google-Proof Q&A

84.2%Weighted 7%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
Artificial Analysis Intelligence IndexReported
29.3%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.
AA-GPQA DiamondReported

Artificial Analysis GPQA Diamond

84.5%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

19.7%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-Omniscience IndexReported

Artificial Analysis Omniscience Index

-46.4%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

20.5%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

84.0%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.
Multilingual1 benchmark
MMLU-ProXProvider exact
81%Weighted 100%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
Multimodal5 benchmarks
MMMUReported

Massive Multi-discipline Multimodal Understanding

81.4%Display only
Source: Reported upstream sourceProvenance: Reported row carried from an upstream public source. Displayable on BenchLM, but not treated as verified unless explicitly marked otherwise.
MMVUReported

Multimodal Multi-disciplinary Video Understanding

72.3%Display only
Source: Reported upstream sourceProvenance: Reported row carried from an upstream public source. Displayable on BenchLM, but not treated as verified unless explicitly marked otherwise.
MathVisionProvider exact
83.9%Display only
Source: Qwen3.5-122B-A10B model cardProvenance: Provider exact
V*Provider exact
92.7%Display only
Source: Qwen3.5-122B-A10B model cardProvenance: Qwen reports V* as with-CI / without-CI. BenchLM stores the first published value.
AA-MMMU-ProReported

Artificial Analysis MMMU-Pro

72.7%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.
Inst. Following2 benchmarks
IFEvalProvider exact

Instruction-Following Eval

91.9%Weighted 35%
Source: Qwen3.5-35B-A3B model cardProvenance: Provider exact
AA-IFBenchReported

Artificial Analysis IFBench

72.5%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 Qwen3.5-35B-A3B perform overall in AI benchmarks?

Qwen3.5-35B-A3B currently ranks #72 out of 200 models on BenchLM's provisional leaderboard with an overall score of 56.97. It also ranks #46 out of 99 on the verified leaderboard. It is created by Alibaba. Its published context window is 262K.

Is Qwen3.5-35B-A3B good for knowledge and understanding?

Qwen3.5-35B-A3B ranks #18 out of 52 models in knowledge and understanding benchmarks with an average score of 75.3. There are stronger options in this category.

Is Qwen3.5-35B-A3B good for coding and programming?

Qwen3.5-35B-A3B ranks #63 out of 122 models in coding and programming benchmarks with an average score of 50.1. There are stronger options in this category.

Is Qwen3.5-35B-A3B good for reasoning and logic?

Qwen3.5-35B-A3B has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.

Is Qwen3.5-35B-A3B good for agentic tool use and computer tasks?

Qwen3.5-35B-A3B ranks #74 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 45.3. There are stronger options in this category.

Is Qwen3.5-35B-A3B good for multimodal and grounded tasks?

Qwen3.5-35B-A3B ranks #29 out of 29 models in multimodal and grounded tasks benchmarks with an average score of 0. There are stronger options in this category.

Is Qwen3.5-35B-A3B good for instruction following?

Qwen3.5-35B-A3B ranks #16 out of 31 models in instruction following benchmarks with an average score of 84.5. There are stronger options in this category.

Is Qwen3.5-35B-A3B good for multilingual tasks?

Qwen3.5-35B-A3B ranks #11 out of 13 models in multilingual tasks benchmarks with an average score of 21.1. There are stronger options in this category.

Is Qwen3.5-35B-A3B open source?

Yes, Qwen3.5-35B-A3B is an open weight model created by Alibaba, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

Does Qwen3.5-35B-A3B have full benchmark coverage on BenchLM?

Not yet. Qwen3.5-35B-A3B currently has 28 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 Qwen3.5-35B-A3B?

Qwen3.5-35B-A3B has a published context window of 262K, which determines how much text it can process in a single interaction.

Last updated: July 23, 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.