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 DirectoryCompare Models

Scoring & Data

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

CharXiv-R Leaderboard

CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.

Updated Aug 17, 2026

Models51
Model coverage51
MetricScore
EvidenceB

On this page

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

CharXiv-R Ranking

Higher score ranks better on this benchmark.

30 of 51 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelANClaude Mythos PreviewAnthropicScore93.2%Percentile100.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank02ModelMAKimi K3Moonshot AIScore91.3%Percentile98.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank03ModelANClaude Opus 4.7AnthropicScore91.0%Percentile96.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank04ModelACQwen3.8-27BAlibaba Cloud / Qwen TeamScore90.2%Percentile94.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank05ModelANClaude Opus 4.8AnthropicScore89.9%Percentile92.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank06ModelGOGemini 3.6 FlashGoogleScore89.4%Percentile90.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank07ModelGOGemini 3.7 FlashGoogleScore88.7%Percentile88.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank08ModelMEMuse Spark 1.1MetaScore88.4%Percentile86.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank09ModelANClaude Sonnet 5AnthropicScore88.3%Percentile84.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank10ModelMAKimi K2.6Moonshot AIScore86.7%Percentile82.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank11ModelMEMuse SparkMetaScore86.4%Percentile80.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank12ModelBYSeed 2.1 ProByteDanceScore86.4%Percentile78.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank13ModelACQwen3.7-PlusAlibaba Cloud / Qwen TeamScore85.9%Percentile76.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank14ModelGOGemini 3.5 FlashGoogleScore84.2%Percentile74.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank15ModelBYSeed 2.1 TurboByteDanceScore83.6%Percentile72.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank16ModelOPGPT-5.2OpenAIScore82.1%Percentile70.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank17ModelOPGPT-5.5 InstantOpenAIScore81.6%Percentile68.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank18ModelACQwen3.6 PlusAlibaba Cloud / Qwen TeamScore81.5%Percentile66.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank19ModelGOGemini 3 ProGoogleScore81.4%Percentile64.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank20ModelOPGPT-5OpenAIScore81.1%Percentile62.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank21ModelXIMiMo-V2.5XiaomiScore81.0%Percentile60.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank22ModelGOGemini 3 FlashGoogleScore80.3%Percentile58.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank23ModelACQwen3.5-27BAlibaba Cloud / Qwen TeamScore79.5%Percentile56.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank24ModelMEMuse Glimmer-30BMetaScore78.8%Percentile54.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank25ModelOPo3OpenAIScore78.6%Percentile52.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank26ModelACQwen3.6-27BAlibaba Cloud / Qwen TeamScore78.4%Percentile50.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank27ModelACQwen3.6-35B-A3BAlibaba Cloud / Qwen TeamScore78.0%Percentile48.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank28ModelMAKimi K2.5Moonshot AIScore77.5%Percentile46.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank29ModelACQwen3.5-35B-A3BAlibaba Cloud / Qwen TeamScore77.5%Percentile44.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank30ModelANClaude Opus 4.6AnthropicScore77.4%Percentile42.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank31ModelTMInkling-SmallThinking Machines LabScore77.4%Percentile40.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank32ModelACQwen3.5-122B-A10BAlibaba Cloud / Qwen TeamScore77.2%Percentile38.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank33ModelGOGemini 3.5 Flash-LiteGoogleScore76.5%Percentile36.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank34ModelGOGemini 3.1 Flash-LiteGoogleScore73.2%Percentile34.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank35ModelOPo4-miniOpenAIScore72.0%Percentile32.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank36ModelACQwen3 VL 235B A22B ThinkingAlibaba Cloud / Qwen TeamScore66.1%Percentile30.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank37ModelACQwen3 VL 32B ThinkingAlibaba Cloud / Qwen TeamScore65.2%Percentile28.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank38ModelACQwen3 VL 32B InstructAlibaba Cloud / Qwen TeamScore62.8%Percentile26.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank39ModelACQwen3 VL 235B A22B InstructAlibaba Cloud / Qwen TeamScore62.1%Percentile24.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank40ModelOPGPT-4oOpenAIScore58.8%Percentile22.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank41ModelOPGPT-4.1 miniOpenAIScore56.8%Percentile20.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank42ModelOPGPT-4.1OpenAIScore56.7%Percentile18.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank43ModelACQwen3 VL 30B A3B ThinkingAlibaba Cloud / Qwen TeamScore56.6%Percentile16.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank44ModelOPGPT-4.5OpenAIScore55.4%Percentile14.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank45ModelACQwen3 VL 8B ThinkingAlibaba Cloud / Qwen TeamScore53.0%Percentile12.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank46ModelCOCommand A+CohereScore52.7%Percentile10.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank47ModelACQwen3 VL 4B ThinkingAlibaba Cloud / Qwen TeamScore50.3%Percentile8.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank48ModelACQwen3 VL 30B A3B InstructAlibaba Cloud / Qwen TeamScore48.9%Percentile6.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank49ModelACQwen3 VL 8B InstructAlibaba Cloud / Qwen TeamScore46.4%Percentile4.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank50ModelOPGPT-4.1 nanoOpenAIScore40.5%Percentile2.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank51ModelACQwen3 VL 4B InstructAlibaba Cloud / Qwen TeamScore39.7%Percentile0.0%Participants51EvidenceCEvaluatedAug 17, 2026

CharXiv-R Highlights

The leading models and scores on this benchmark.

Rank #1Claude Mythos Preview93.2%Rank #2Kimi K391.3%Rank #3Claude Opus 4.791.0%Rank #4Qwen3.8-27B90.2%

CharXiv-R Score Distribution

A closer view of the leading scores on this benchmark.

CharXiv-R

The Top AI Models for CharXiv-R

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

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

  1. 01
    AN
    Claude Mythos PreviewAnthropic
    Score
    93.2%

    Strengths

    • Ranks #1 of 51 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures CharXiv-R, not total model capability
  2. 02
    MA
    Kimi K3Moonshot AI
    Score
    91.3%
    Price
    $3.0 input / $15 output per 1M tokens
    Speed
    Up to 26 tok/s via Fireworks

    Strengths

    • Ranks #2 of 51 compared models
    • 98th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures CharXiv-R, not total model capability
  3. 03
    AN
    Claude Opus 4.7Anthropic
    Score
    91.0%
    Price
    $5.0 input / $25 output per 1M tokens
    Speed
    Up to 42 tok/s via Anthropic

    Strengths

    • Ranks #3 of 51 compared models
    • 96th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures CharXiv-R, not total model capability
  4. 04
    AC
    Qwen3.8-27BAlibaba Cloud / Qwen Team
    Score
    90.2%

    Strengths

    • Ranks #4 of 51 compared models
    • 94th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures CharXiv-R, not total model capability
  5. 05
    AN
    Claude Opus 4.8Anthropic
    Score
    89.9%
    Price
    $5.0 input / $25 output per 1M tokens
    Speed
    Up to 46 tok/s via Anthropic

    Strengths

    • Ranks #5 of 51 compared models
    • 92th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures CharXiv-R, not total model capability

Selection summary

Best AI Models for CharXiv-R

Claude Mythos Preview currently leads CharXiv-R with 93.2%. 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.

Benchmark rank #1Claude Mythos Preview93.2%Benchmark rank #2Kimi K391.3% · $3.0 input / $15 output per 1M tokensBenchmark rank #3Claude Opus 4.791.0% · $5.0 input / $25 output per 1M tokens

What is CharXiv-R?

What CharXiv-R measures and how its scores work.

CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.

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

Family
CharXiv-R
Modality
multimodal
Primary category
multimodal
Score direction
higher
LLMBoard eligible
Yes
Evaluation key
charxiv-r|llm-stats-current

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

FAQ

Common questions about CharXiv-R.

Which model scores highest on CharXiv-R?

Claude Mythos Preview is currently ranked first with 93.2%.

What does CharXiv-R measure?

CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

51 model results are currently shown.

Does this benchmark affect the overall score?

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