llmboard.aiLeaderboard Center
Overall
Overall RankingOpen Models
Tools
Model DirectoryCompare Models
Capabilities
CodingReasoningMathKnowledgeInstruction Following
Price & Efficiency
Price & ValueCapability vs. PriceRuntime Performance
Modalities
Image GenerationVideo GenerationSpeech ModelsEmbeddings
Core Benchmarks
GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-ProView all benchmarks
Methods
Scoring & Data
393 models668 benchmarks

Leaderboard Center

Overall RankingCodingCore BenchmarksPrice & ValueRuntime Performance

Modalities

All ModelsImage GenerationImage EditingVideo GenerationImage-to-VideoVideo EditingText-to-SpeechSpeech-to-TextEmbeddings

Data & Methods

Scoring MethodAll BenchmarksReasoningMath

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai

multimodal benchmark

OSWorld-Verified

OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.

Updated Aug 11, 2026

Models23
Model coverage23
MetricScore
EvidenceB

On this page

  • Ranking
  • Distribution
  • Highlights
  • About
  • FAQ

OSWorld-Verified Ranking

Higher score ranks better on this benchmark.

23 rows
Columns

Show columns

01ACQwen3.8 MaxAlibaba Cloud / Qwen Team86.1%100.0%23CAug 11, 2026
02ANClaude Fable 5Anthropic85.0%95.5%23CAug 11, 2026
03ANClaude Opus 4.8Anthropic83.4%90.9%23CAug 11, 2026
04GOGemini 3.6 FlashGoogle83.0%86.4%23CAug 11, 2026
05ANClaude Sonnet 5Anthropic81.2%81.8%23CAug 11, 2026
06MEMuse Spark 1.1Meta80.8%77.3%23CAug 11, 2026
07ANClaude Mythos PreviewAnthropic79.6%72.7%23CAug 11, 2026
08OPGPT-5.5OpenAI78.7%68.2%23CAug 11, 2026
09GOGemini 3.5 FlashGoogle78.4%63.6%23CAug 11, 2026
10ANClaude Opus 4.7Anthropic78.0%59.1%23CAug 11, 2026
11OPGPT-5.4OpenAI75.0%54.5%23CAug 11, 2026
12GOGemini 3.5 Flash-LiteGoogle74.0%50.0%23CAug 11, 2026
13ACQwen3.7-PlusAlibaba Cloud / Qwen Team73.3%45.5%23CAug 11, 2026
14MAKimi K2.6Moonshot AI73.1%40.9%23CAug 11, 2026
15OPGPT-5.4 miniOpenAI72.1%36.4%23CAug 11, 2026
16MIMiniMax M3MiniMax70.1%31.8%23CAug 11, 2026
17MEMuse Glimmer-30BMeta65.9%27.3%23CAug 11, 2026
18OPGPT-5.3 CodexOpenAI64.7%22.7%23CAug 11, 2026
19ACQwen3.6 PlusAlibaba Cloud / Qwen Team62.5%18.2%23CAug 11, 2026
20ACQwen3.5-122B-A10BAlibaba Cloud / Qwen Team58.0%13.6%23CAug 11, 2026
21ACQwen3.5-27BAlibaba Cloud / Qwen Team56.2%9.1%23CAug 11, 2026
22ACQwen3.5-35B-A3BAlibaba Cloud / Qwen Team54.5%4.5%23CAug 11, 2026
23OPGPT-5.4 nanoOpenAI39.0%0.0%23CAug 11, 2026

OSWorld-Verified Score Distribution

A closer view of the leading scores on this benchmark.

OSWorld-Verified

OSWorld-Verified Highlights

The leading models and scores on this benchmark.

Rank #1Qwen3.8 Max86.1%Rank #2Claude Fable 585.0%Rank #3Claude Opus 4.883.4%Rank #4Gemini 3.6 Flash83.0%

What is OSWorld-Verified?

What OSWorld-Verified measures and how its scores work.

OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.

Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.

Family
OSWorld-Verified
Modality
multimodal
Primary category
multimodal
Score direction
higher
LLMBoard eligible
Yes
Evaluation key
osworld-verified|llm-stats-current

Benchmark scores retain their original unit. Overall score eligibility is shown separately.

FAQ

Common questions about OSWorld-Verified.

Which model scores highest on OSWorld-Verified?

Qwen3.8 Max is currently ranked first with 86.1%.

What does OSWorld-Verified measure?

OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

23 model results are currently shown.

Does this benchmark affect the overall score?

Yes. This benchmark can contribute to the current LLMBoard capability score.