llmboard.aiAI model intelligence
Home

Model Rankings

OverallOpen ModelsAgentCodingReasoningMathKnowledgeInstruction FollowingTextVision
Image GenerationImage Editing
Video GenerationImage to VideoVideo Editing
Text to SpeechSpeech to Text
Embeddings

Efficiency

Chat Token PricingImage PricingVideo PricingAudio Pricing
Chat Speed & LatencyProvider Reliability

Benchmarks

GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-Pro
All Benchmarks

Tools

Model Directory

Scoring & Data

Scoring & Data
1438 models809 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 Benchmarks
llmboard.aiCopyright 2026 llmboard.ai

reasoning benchmark

BrowseComp Leaderboard

BrowseComp is a reasoning benchmark comprising 1,266 questions that challenge AI agents to navigate the internet to find hard-to-find, entangled information.

Updated Nov 13, 2026

Models77
Model coverage77
MetricScore
EvidenceB

On this page

  • Ranking
  • Highlights
  • Distribution
  • Top models
  • About
  • FAQ

BrowseComp Ranking

Higher score ranks better on this benchmark.

30 of 77 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelSAAtria Dawn PreviewShanghai AI LaboratoryScore92.50%Percentile100.00%Participants77EvidenceCEvaluatedOct 12, 2026
Rank02ModelMAKimi K3.1 ProMoonshot AIScore92.10%Percentile98.68%Participants77EvidenceCEvaluatedNov 12, 2026
Rank03ModelMELongCat-2.5MeituanScore91.90%Percentile97.37%Participants77EvidenceCEvaluatedNov 12, 2026
Rank04ModelOPGPT-6 AstraOpenAIScore91.50%Percentile96.05%Participants77EvidenceCEvaluatedNov 12, 2026
Rank05ModelMAKimi K3Moonshot AIScore91.20%Percentile94.74%Participants77EvidenceCEvaluatedNov 12, 2026
Rank06ModelANClaude Opus 5AnthropicScore90.80%Percentile93.42%Participants77EvidenceCEvaluatedNov 12, 2026
Rank07ModelOPGPT-5.6 SolOpenAIScore90.40%Percentile92.11%Participants77EvidenceCEvaluatedNov 12, 2026
Rank08ModelTMInkling 2 SmallThinking Machines LabScore90.20%Percentile90.79%Participants77EvidenceCEvaluatedNov 12, 2026
Rank09ModelOPGPT-5.5 ProOpenAIScore90.10%Percentile89.47%Participants77EvidenceCEvaluatedNov 12, 2026
Rank10ModelOPGPT-5.6 TerraOpenAIScore87.50%Percentile88.16%Participants77EvidenceCEvaluatedNov 12, 2026
Rank11ModelBAERNIE 5.5 Flash PreviewBaiduScore87.40%Percentile86.84%Participants77EvidenceCEvaluatedNov 12, 2026
Rank12ModelANClaude Mythos PreviewAnthropicScore86.90%Percentile85.53%Participants77EvidenceCEvaluatedNov 12, 2026
Rank13ModelMAKimi K2.6Moonshot AIScore86.30%Percentile84.21%Participants77EvidenceCEvaluatedNov 12, 2026
Rank14ModelBYSeed 2.1 ProByteDanceScore86.20%Percentile82.89%Participants77EvidenceCEvaluatedNov 12, 2026
Rank15ModelGOGemini 3.1 ProGoogleScore85.90%Percentile81.58%Participants77EvidenceCEvaluatedNov 12, 2026
Rank16ModelBYSeed 2.1 TurboByteDanceScore84.90%Percentile80.26%Participants77EvidenceCEvaluatedNov 12, 2026
Rank17ModelANClaude Sonnet 5AnthropicScore84.70%Percentile78.95%Participants77EvidenceCEvaluatedNov 12, 2026
Rank18ModelMAMistral Large 4Mistral AIScore84.50%Percentile77.63%Participants77EvidenceCEvaluatedNov 12, 2026
Rank19ModelOPGPT-5.5OpenAIScore84.40%Percentile76.32%Participants77EvidenceCEvaluatedNov 12, 2026
Rank20ModelANClaude Opus 4.8AnthropicScore84.30%Percentile75.00%Participants77EvidenceCEvaluatedNov 12, 2026
Rank21ModelTEHy3TencentScore84.20%Percentile73.68%Participants77EvidenceCEvaluatedNov 12, 2026
Rank22ModelANClaude Opus 4.6AnthropicScore84.00%Percentile72.37%Participants77EvidenceCEvaluatedNov 12, 2026
Rank23ModelMIMiniMax M3MiniMaxScore83.52%Percentile71.05%Participants77EvidenceCEvaluatedNov 12, 2026
Rank24ModelDEDeepSeek-V4-Pro-MaxDeepSeekScore83.40%Percentile69.74%Participants77EvidenceCEvaluatedNov 12, 2026
Rank25ModelOPGPT-5.6 LunaOpenAIScore83.30%Percentile68.42%Participants77EvidenceCEvaluatedNov 12, 2026
Rank26ModelOPGPT-5.4OpenAIScore82.70%Percentile67.11%Participants77EvidenceCEvaluatedNov 12, 2026
Rank27ModelANClaude Opus 4.7AnthropicScore79.30%Percentile65.79%Participants77EvidenceCEvaluatedNov 12, 2026
Rank28ModelZAGLM-5.1Zhipu AIScore79.30%Percentile64.47%Participants77EvidenceCEvaluatedNov 12, 2026
Rank29ModelNVNemotron 3.5 Super VL (120B A12B)NVIDIAScore78.00%Percentile63.16%Participants77EvidenceCEvaluatedNov 12, 2026
Rank30ModelOPGPT-5.2 ProOpenAIScore77.90%Percentile61.84%Participants77EvidenceCEvaluatedNov 12, 2026

BrowseComp Highlights

The leading models and scores on this benchmark.

BrowseComp Score Distribution

A closer view of the leading scores on this benchmark.

BrowseComp

The Top AI Models for BrowseComp

The first five results on this benchmark, with official price and output speed added where the model identity can be matched.

What is BrowseComp?

What BrowseComp measures and how its scores work.

BrowseComp is a benchmark of 1,266 questions requiring AI agents to search the internet for concise, verifiable answers.

It measures agents' persistence in information gathering, creativity in web navigation, and ability to find concise, verifiable answers.

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

Family
BrowseComp
Modality
text
Primary category
reasoning
Score direction
higher
LLMBoard eligible
No
Evaluation key
overall

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

FAQ

Common questions about BrowseComp.

Which model scores highest on BrowseComp?

Atria Dawn Preview is currently ranked first with 92.50%.

What are the top three models on BrowseComp?

The current leaders are Atria Dawn Preview (92.50%), Kimi K3.1 Pro (92.10%), and LongCat-2.5 (91.90%).

Which BrowseComp model has the lowest official input price?

Step-3.5-Flash has the lowest matched official input price at $0.10 input / $0.30 output per 1M tokens.

Which models are fastest among BrowseComp results?

The fastest matched records are MiniMax M3 (156.78 tok/s via MiniMax), Step-3.5-Flash (150.00 tok/s via StepFun), and Kimi K3.1 Pro (142.49 tok/s via Moonshot AI).

Does the highest score result prove overall model quality?

No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.

What does BrowseComp measure?

It measures agents' persistence in information gathering, creativity in web navigation, and ability to find concise, verifiable answers.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

77 model results are currently shown.

Does this benchmark affect the overall score?

No. This benchmark is shown for reference but does not contribute to the overall score.

Rank #1Atria Dawn Preview92.50%
Rank #2Kimi K3.1 Pro92.10%
Rank #3LongCat-2.591.90%
Rank #4GPT-6 Astra91.50%

Ranking basisThis browsecomp AI model leaderboard uses descending score in the benchmark original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.

  1. 01
    SA
    Atria Dawn PreviewShanghai AI Laboratory
    Score
    92.50%

    Strengths

    • Ranks #1 of 77 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures BrowseComp, not total model capability
  2. 02
    MA
    Kimi K3.1 ProMoonshot AI
    Score
    92.10%
    Price
    $2.0 input / $10 output per 1M tokens
    Speed
    Up to 142.49 tok/s via Moonshot AI

    Strengths

    • Ranks #2 of 77 compared models
    • 99th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures BrowseComp, not total model capability
  3. 03
    ME
    Meituan
    Score
    91.90%

    Strengths

    • Ranks #3 of 77 compared models
    • 97th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures BrowseComp, not total model capability
  4. 04
    OP
    OpenAI
    Score
    91.50%
    Price
    $10 input / $50 output per 1M tokens
    Speed
    Up to 86.85 tok/s via OpenAI

    Strengths

    • Ranks #4 of 77 compared models
    • 96th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures BrowseComp, not total model capability
  5. 05
    MA
    Moonshot AI
    Score
    91.20%
    Price
    $3.0 input / $15 output per 1M tokens

    Strengths

    • Ranks #5 of 77 compared models
    • 95th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures BrowseComp, not total model capability

Selection summary

Best AI Models for BrowseComp

Atria Dawn Preview currently leads BrowseComp with 92.50%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price, and runtime.

Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.

LongCat-2.5
GPT-6 Astra
Kimi K3
Benchmark rank #1Atria Dawn Preview92.50%
Benchmark rank #2Kimi K3.1 Pro92.10% · $2.0 input / $10 output per 1M tokens
Benchmark rank #3LongCat-2.591.90%