Skip to main content
Radar

Five or fewer confirmed AI changes, with original sources, on mornings when something changed.A free source-linked morning brief.

Start the free Radar Brief

Qwen3.6-35B-A3B

CurrentReleased Apr 15, 2026Open WeightReasoning262K context

Released Apr 15, 2026 see all recent releases

Decision reading
Qwen3.6-35B-A3B scores 51.6 out of 100 and ranks #122 of 228. This profile shows 41 source-displayable benchmark rows; its strongest eligible category is Multimodal & Grounded at #26. No comparable first-party API price is published in the catalog.

Data as of August 29, 2026 · How the score is built

Strongest published evidence

Multimodal & Grounded ranks #26. Particularly strong for screenshots, documents, charts, and grounded multimodal workflows.

Validate before choosing

41 published rows leave some tracked benchmark slots empty. No comparable first-party API token rate is published.

Decision snapshot

Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.

Capability

51.6/100

field median 58.8

#122 of 228 ranked models

Price

Not listed

input median $1

No comparable first-party hosted token rate

Speed

124tok/s

field median 86 tok/s

First token 45.65 s

Context

262Ktokens

field median 256,000

Reported for this model; direct source link not stored

Capability shape

Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.

Qwen3.6-35B-A3B category percentile values

  • Agentic37th percentile
  • Coding48th percentile
  • ReasoningNot eligible
  • Knowledge3rd percentile
  • MathNot eligible
  • MultilingualNot eligible
  • Multimodal26th percentile
  • Instruction followingNot eligible

The dashed outline is median of 6 nearest peers.

Eligible category ranks

  1. Agentic#88/140
  2. Coding#76/146
  3. ReasoningNot ranked
  4. Knowledge#59/61
  5. MathNot ranked
  6. MultilingualNot ranked
  7. Multimodal#26/35
  8. Inst. FollowingNot ranked
Top decileTop quartileMid-fieldNot eligible

How much of this is verified

Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.

  1. Agentic11/11 verified
  2. Coding6/6 verified
  3. ReasoningNot measured
  4. Knowledge5/5 verified
  5. Math5/5 verified
  6. MultilingualNot measured
  7. Multimodal14/14 verified
  8. Inst. FollowingNot measured
Verified sourceProvisionalNot measured

Spec sheet

Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.

API model ID
Not published
Context window
262K
Maximum output
Not sourced yet
Knowledge cutoff
Not sourced yet
Input modalities
Not sourced yet
Output modalities
Not sourced yet
Parameters
Not sourced yet
Availability
Not sourced yet
Cloud regions
Not tracked yet
Lifecycle
Current
API capabilities
Tool calling, structured outputs, and batch support are not tracked yet
Prompt caching
Not documented in the pricing record
Self-host
Open weights available; hardware estimate not sourced
Rate limits
Not tracked yet

Deployment options

Self-host and provider-specific paths stay separate from benchmark evidence so operating constraints are visible before a score becomes the whole decision.

Published weights are available, but BenchLM does not yet have a sourced parameter and VRAM profile for this exact model. Hardware cost estimates stay unavailable until that sizing record is complete.

Estimate VRAM from known parameters

Category score record

Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #88 of 140Percentile 37thWeight 22%11 benchmarksVerified45.2
CodingRank #76 of 146Percentile 48thWeight 20%6 benchmarksVerified50.5
ReasoningWeight 17%0 benchmarksNot measuredNot measured
KnowledgeRank #59 of 61Percentile 3rdWeight 12%5 benchmarksVerified42.8
MathRank Not rankedWeight 5%5 benchmarksVerified71.6
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalRank #26 of 35Percentile 26thWeight 12%14 benchmarksVerified42.4
Inst. FollowingWeight 5%0 benchmarksNot measuredNot measured

Benchmark ledger

Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.

Coding6 rows
Coding benchmark values, best verified comparison, weight, and source status
LiveCodeBenchLiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for CodeScore80.4%Versus best verified row

Best verified: Qwen3.7 Max · 91.6%

Gap11.2 behindWeightWeighted 38%
SWE-bench VerifiedSoftware Engineering Benchmark VerifiedScore73.4%Versus best verified row

Best verified: Claude Opus 5 · 96%

Gap22.6 behindWeightWeighted 16%
SWE-bench ProScore49.5%Versus best verified row

Best verified: Claude Mythos 5 · 80.3%

Gap30.8 behindWeightWeighted 10%
SWE MultilingualScore67.2%Versus best verified row

Best verified: Claude Opus 5 · 89.5%

Gap22.3 behindWeightDisplay only
Terminal-Bench 2.0Score51.5%Versus best verified row

Best verified: GPT-5.6 Sol · 91.9%

Gap40.4 behindWeightDisplay only
NL2RepoScore29.4%Versus best verified row

Best verified: DeepSeek V4 Pro 0813 · 61.5%

Gap32.1 behindWeightDisplay only
Agentic11 rows
Agentic benchmark values, best verified comparison, weight, and source status
Terminal-Bench 2.0Score51.5%Versus best verified row

Best verified: GPT-5.6 Sol · 91.9%

Gap40.4 behindWeightWeighted 38%
Claw-EvalScore68.7%Versus best verified row

Best verified: Ornith-1.5-397B · 81.4%

Gap12.7 behindWeightDisplay only
QwenClawBenchScore52.6%Versus best verified row

Best verified: Qwen3.7 Max · 64.3%

Gap11.7 behindWeightDisplay only
QwenWebBenchScore1397Versus best verified row

Best verified: Qwen3.7 Max · 1568

Gap171 behindWeightDisplay only
τ³-bench resultsτ³-Bench Tool-Agent-User EvaluationScore67.2%Versus best verified row

Best verified: Mistral Medium 3.5 128B · 91.4%

Gap24.2 behindWeightDisplay only
VITA-BenchScore35.6%Versus best verified row

Best verified: Qwen3.7 Max · 47.9%

Gap12.3 behindWeightDisplay only
DeepPlanningScore25.9%Versus best verified row

Best verified: Qwen3.7 Plus · 62.3%

Gap36.4 behindWeightDisplay only
ToolathlonScore26.9%Versus best verified row

Best verified: Muse Spark 1.1 · 75.6%

Gap48.7 behindWeightDisplay only
MCP AtlasScore62.8%Versus best verified row

Best verified: Muse Spark 1.1 · 88.1%

Gap25.3 behindWeightDisplay only
WideResearchScore60.1%Versus best verified row

Best verified: Hy4 preview · 83.9%

Gap23.8 behindWeightDisplay only
Gert LabsGert Labs Composite Game BenchmarkScore42.65%Versus best verified row

Best verified: Claude Opus 4.8 · 72.97%

Gap30.3 behindWeightDisplay only
Benchmark exact
Knowledge5 rows
Knowledge benchmark values, best verified comparison, weight, and source status
HLEHumanity's Last ExamScore21.4%Versus best verified row

Best verified: Claude Opus 5 · 64.7%

Gap43.3 behindWeightWeighted 45%
MMLU-ProMassive Multitask Language Understanding ProfessionalScore85.2%Versus best verified row

Best verified: Qwen3.7 Max · 89.6%

Gap4.4 behindWeightWeighted 30%
SuperGPQASuperGPQA: Scaling LLM Evaluation Across 285 Graduate DisciplinesScore64.7%Versus best verified row

Best verified: Qwen 3.6 Max (preview) · 73.9%

Gap9.2 behindWeightWeighted 7%
GPQAGraduate-Level Google-Proof Q&AScore86%Versus best verified row

Best verified: Sakana Fugu-Ultra · 95.5%

Gap9.5 behindWeightWeighted 7%
C-EvalScore90%Versus best verified row

Best verified: Qwen3.6 Plus · 93.3%

Gap3.3 behindWeightDisplay only
Math5 rows
Math benchmark values, best verified comparison, weight, and source status
HMMT Feb 2026Harvard-MIT Mathematics Tournament February 2026Score83.6%Versus best verified row

Best verified: Qwen3.7 Max · 97.1%

Gap13.5 behindWeightWeighted 25%
AIME26AIME 2026Score92.7%Versus best verified row

Best verified: GLM-5.2 · 99.2%

Gap6.5 behindWeightWeighted 25%
HMMT Feb 2025Harvard-MIT Mathematics Tournament February 2025Score90.7%Versus best verified row

Best verified: Qwen3.6 Plus · 96.7%

Gap6 behindWeightDisplay only
HMMT Nov 2025Harvard-MIT Mathematics Tournament November 2025Score89.1%Versus best verified row

Best verified: Qwen3.6 Plus · 94.6%

Gap5.5 behindWeightDisplay only
MMAnswerBenchScore78.9%Versus best verified row

Best verified: GLM-5.2 · 91.0%

Gap12.1 behindWeightDisplay only
Multimodal14 rows
Multimodal benchmark values, best verified comparison, weight, and source status
MMMU-ProMassive Multi-discipline Multimodal Understanding ProScore75.3%Versus best verified row

Best verified: GPT-5.4 Pro · 94%

Gap18.7 behindWeightWeighted 45%
CharXivCharXiv ReasoningScore78%Versus best verified row

Best verified: Claude Mythos 5 · 93.5%

Gap15.5 behindWeightWeighted 25%
MMMUMassive Multi-discipline Multimodal UnderstandingScore81.7%Versus best verified row

Best verified: Qwen3.6-27B · 82.9%

Gap1.2 behindWeightDisplay only
RealWorldQAScore85.3%Versus best verified row

Best verified: Qwen3.8-Flash-Next · 88.5%

Gap3.2 behindWeightDisplay only
OmniDocBench 1.5Score89.9%Versus best verified row

Best verified: Qwen3.8 Max · 92.1%

Gap2.2 behindWeightDisplay only
SimpleVQAScore58.9%Versus best verified row

Best verified: Qwen3.7 Plus · 81.7%

Gap22.8 behindWeightDisplay only
CC-OCRScore81.9%Versus best verified row

Best verified: Qwen3.6-35B-A3B · 81.9%

GapBest verifiedWeightDisplay only
AI2D_TESTAI2D test splitScore92.7%Versus best verified row

Best verified: Qwen3.6-35B-A3B · 92.7%

GapBest verifiedWeightDisplay only
RefCOCO (avg)RefCOCO averageScore92.0%Versus best verified row

Best verified: Qwen3.6-27B · 92.5%

Gap0.5 behindWeightDisplay only
ODINW13Score50.8%Versus best verified row

Best verified: Qwen3.7 Plus · 51.1%

Gap0.3 behindWeightDisplay only
Video-MME (with subtitle)Video-MME with subtitleScore86.6%Versus best verified row

Best verified: Qwen3.8 Max · 90.4%

Gap3.8 behindWeightDisplay only
Video-MME (w/o subtitle)Video-MME without subtitleScore82.5%Versus best verified row

Best verified: Qwen3.6-35B-A3B · 82.5%

GapBest verifiedWeightDisplay only
VideoMMMUScore83.7%Versus best verified row

Best verified: Qwen3.8 Max · 88.7%

Gap5 behindWeightDisplay only
MLVU (M-Avg)MLVU mean averageScore86.2%Versus best verified row

Best verified: Qwen3.8 Max · 90.8%

Gap4.6 behindWeightDisplay only

Lineage

The sequence follows explicit supersedes links. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.

  1. Apr 15, 2026 · you are here

    Qwen3.6-35B-A3B

    Score 51.5 · Price not listed

Base entry

How to read this profile

The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.

Qwen3.6-35B-A3B ranks #122 of 228 on the public leaderboard with a score of 51.55/100. It does not yet have enough sourced coverage for a verified position.

Qwen3.6-35B-A3B is a open weight model with a 262K context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.

Official exact-value snapshot from the Qwen/Qwen3.6-35B-A3B Hugging Face model card. The public repository was created on April 15, 2026, and the model card states a native 262,144-token context window extensible to 1,010,000 tokens.

41 of 408 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.

Its strongest eligible category is Multimodal & Grounded at #26, while its lowest eligible position is Agentic at #88. particularly strong for screenshots, documents, charts, and grounded multimodal workflows.

Radar

Qwen3.6-35B-A3B release history

Full release history

Radar confirmed these at the source. Start the free Radar Brief

Frequently asked questions

How does Qwen3.6-35B-A3B perform overall in AI benchmarks?

Qwen3.6-35B-A3B ranks #122 out of 228 models on the public BenchAlign leaderboard, with a score of 51.55/100. Its evidence status is Estimated, and this profile shows 41 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.

Is Qwen3.6-35B-A3B good for knowledge and understanding?

Qwen3.6-35B-A3B ranks #59 out of 61 eligible models for knowledge and understanding, with a public category score of 42.8/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.

Is Qwen3.6-35B-A3B good for coding and programming?

Qwen3.6-35B-A3B ranks #76 out of 146 eligible models for coding and programming, with a public category score of 50.5/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.

Is Qwen3.6-35B-A3B good for mathematics?

Qwen3.6-35B-A3B has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is Qwen3.6-35B-A3B good for agentic tool use and computer tasks?

Qwen3.6-35B-A3B ranks #88 out of 140 eligible models for agentic tool use and computer tasks, with a public category score of 45.2/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.

Is Qwen3.6-35B-A3B good for multimodal and grounded tasks?

Qwen3.6-35B-A3B ranks #26 out of 35 eligible models for multimodal and grounded tasks, with a public category score of 42.4/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.

Is Qwen3.6-35B-A3B open source?

Qwen3.6-35B-A3B is an open-weight model from Alibaba. Its weights can be downloaded for local or hosted deployment, subject to the published license. Open weight does not automatically mean open source: training data and training code may remain private, and commercial restrictions can still apply.

Does Qwen3.6-35B-A3B have full benchmark coverage on BenchLM?

No. Qwen3.6-35B-A3B currently has 57 source-displayable rows across 408 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.

What is the context window size of Qwen3.6-35B-A3B?

Qwen3.6-35B-A3B has a reported context window of 262K in the exact-model catalog record. The value stays visible, but the profile marks its source link as unavailable instead of presenting it as directly documented. Maximum output length remains separate because providers often publish a different limit.

Compare Qwen3.6-35B-A3B with every tracked model399 comparisons

Last updated August 29, 2026. Runtime fields remain blank until a sourced snapshot exists.

Watch Qwen3.6-35B-A3B in the weekly brief

Get one weekly email when material rank, price, availability, or benchmark evidence changes are worth revisiting.

Read a sample issue

Join 2,000+ readers.