Skip to main content
BenchLM
Advertise

Model profile

Claude Opus 4.5

AnthropicEstablishedReleased November 2025
Data verified
Overall Score
64.22Public #34 of 200Verified #28 of 99
Arena Elo
1468
Eligible category ranks
7of 8
Price (1M tokens)
$5 in / $25 out
API pricing
Speed
46tok/s
Context
200K

Evidence coverage

59 of 321 tracked benchmarks are published. 11 are verified and 48 provisional. 8 of 8 categories are measured.

Updated July 20, 2026Methodology
Published / tracked
59 / 321
Verified
11
Provisional
48
Categories with evidence
8 / 8

Evidence by category

  • Agentic16 benchmarks
    Mixed evidence
  • Coding6 benchmarks
    Mixed evidence
  • Reasoning4 benchmarks
    Reported
  • Knowledge13 benchmarks
    Mixed evidence
  • Math7 benchmarks
    Mixed evidence
  • Multilingual2 benchmarks
    Reported
  • Multimodal8 benchmarks
    Reported
  • Inst. Following3 benchmarks
    Reported
ProprietaryNon-Reasoning
Confidence:
Very high
base

Claude Opus 4.5 ranks #34 out of 200 models on the public leaderboard with an overall score of 64.22/100. It also ranks #28 out of 99 on the verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

Claude Opus 4.5 is a proprietary model with a 200K token context window. It processes queries without explicit chain-of-thought reasoning, offering faster response times and lower token usage.

Claude Opus 4.5 sits inside the Claude Opus 4.5 family alongside Claude Opus 4.5 Thinking. This profile currently has 59 of 321 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 (#2), while its weakest is Agentic (#106). 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 63.5–64.75

  1. Gemini 3.5 Flash
    Google
    #3364.75
    Gemini 3.5 Flash is #33 with a score of 64.75.
    Compare
  2. Claude Opus 4.5Current model
    Anthropic
    #3464.22
    Claude Opus 4.5 is #34 with a score of 64.22.
  3. Claude Opus 4.6 (Adaptive)
    Anthropic
    #3564.18
    Claude Opus 4.6 (Adaptive) is #35 with a score of 64.18.
    Compare
  4. MiniMax M2.7
    MiniMax
    #3664.11
    MiniMax M2.7 is #36 with a score of 64.11.
    Compare
  5. GLM-5.2
    Z.AI
    #3763.96
    GLM-5.2 is #37 with a score of 63.96.
    Compare
  6. GPT-5.5 Pro
    OpenAI
    #3863.69
    GPT-5.5 Pro is #38 with a score of 63.69.
    Compare
  7. GLM-5V-Turbo
    Z.AI
    #3963.5
    GLM-5V-Turbo is #39 with a score of 63.5.
    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. Multilingual92%
    Eligible cohort rank #2 of 13Category score 82.9
  2. Math17%
    Eligible cohort rank #6 of 7Category score 58.6
  3. Coding84%
    Eligible cohort rank #20 of 122Category score 61.9
  4. Multimodal11%
    Eligible cohort rank #26 of 29Category score 28.4
  5. Inst. Following22%
    Eligible cohort rank #30 of 38Category score 56.4
  6. Knowledge27%
    Eligible cohort rank #38 of 52Category score 62.2
  7. Agentic11%
    Eligible cohort rank #106 of 119Category score 35.9

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 #106 of 119Percentile 11thWeight 22%16 benchmarksMixed sources35.9
CodingRank #20 of 122Percentile 84thWeight 20%6 benchmarksMixed sources61.9
ReasoningRank Not rankedWeight 17%4 benchmarksReported91.5
KnowledgeRank #38 of 52Percentile 27thWeight 12%13 benchmarksMixed sources62.2
MathRank #6 of 7Percentile 17thWeight 5%7 benchmarksMixed sources58.6
MultilingualRank #2 of 13Percentile 92ndWeight 7%2 benchmarksReported82.9
MultimodalRank #26 of 29Percentile 11thWeight 12%8 benchmarksReported28.4
Inst. FollowingRank #30 of 38Percentile 22ndWeight 5%3 benchmarksReported56.4

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 Overall1468Not availableNot available

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.

Agentic16 benchmarks
Terminal-Bench 2.0Provider exact
59.3%Weighted 38%
Source: Anthropic: Claude Opus 4.5 system cardProvenance: Provider exact
OSWorld-VerifiedProvider exact
66.3%Weighted 34%
Source: Anthropic: Claude Opus 4.5 system cardProvenance: Provider exact
OSWorldProvider exact
66.3%Display only
Source: Anthropic: Claude Opus 4.5 system cardProvenance: Provider exact
Claw-EvalSecondary exact
59.6%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
QwenClawBenchSecondary exact
52.3%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
τ³-bench resultsSecondary exact

τ³-Bench Tool-Agent-User Evaluation

70.2%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
VITA-BenchBenchmark exact
23.3%Display only
Source: VitaBench leaderboardProvenance: Benchmark exact
DeepPlanningSecondary exact
26.4%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
ToolathlonSecondary exact
43.5%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
MCP AtlasSecondary exact
42.3%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
MCP-TasksSecondary exact
71.8%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
WideResearchSecondary exact
76.4%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
CyberGymBenchmark exact
50.6%Display only
Source: CyberGym leaderboardProvenance: CyberGym Level 1 reports Anthropic Agent with Claude Opus 4.5 on the public leaderboard. BenchLM stores the reported target-vulnerability reproduction success rate on the local cyberGym key.
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

86.3%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.
Gert LabsBenchmark exact

Gert Labs Composite Game Benchmark

64.23%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.
JobBenchBenchmark exact
32.3%Display only
Source: JobBench paperProvenance: JobBench reports Opus-4.5 under Claude Code on the main-set leaderboard. BenchLM stores the reported main-set score.
Coding6 benchmarks
SWE-bench VerifiedProvider exact

Software Engineering Benchmark Verified

80.9%Weighted 16%
Source: Anthropic: Claude Opus 4.5 system cardProvenance: Provider exact
SWE-bench ProSecondary exact
57.1%Weighted 10%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
LiveCodeBench v6Secondary exact
84.8%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
SWE MultilingualSecondary exact
77.5%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
NL2RepoSecondary exact
43.2%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
AA-SciCodeReported

Artificial Analysis SciCode

47.0%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.
Reasoning4 benchmarks
LongBench v2Secondary exact
64.4%Weighted 38%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
AI-NeedleSecondary exact
74%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
AA-LCRReported

Artificial Analysis Long Context Reasoning

65.3%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.
CritPtReported

Critical Physics Tasks

0.3%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.
Knowledge13 benchmarks
HLESecondary exact

Humanity's Last Exam

30.8%Weighted 45%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
MMLU-ProSecondary exact

Massive Multitask Language Understanding Professional

89.5%Weighted 30%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
GPQAProvider exact

Graduate-Level Google-Proof Q&A

87%Weighted 7%
Source: Anthropic: Claude Opus 4.5 system cardProvenance: Provider exact
SuperGPQASecondary exact

SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines

70.6%Weighted 7%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
MMLU-ReduxSecondary exact
96.6%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
C-EvalSecondary exact
92.2%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
Artificial Analysis Intelligence IndexReported
34.7%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-GPQA DiamondReported

Artificial Analysis GPQA Diamond

81.0%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.
AA-HLEReported

Artificial Analysis Humanity's Last Exam

12.9%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.
AA-Omniscience IndexReported

Artificial Analysis Omniscience Index

-3.9%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.
AA-Omniscience AccuracyReported

Artificial Analysis Omniscience Accuracy

40.7%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 Hallucination RateReported

Artificial Analysis Omniscience Hallucination Rate

75.4%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 MMLU-ProReported

Artificial Analysis MMLU-Pro

88.9%Display only
Source: Artificial Analysis: mmlu-pro leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
Math7 benchmarks
FrontierMath v2 (Tiers 1-3)Benchmark exact

FrontierMath v2 Tiers 1-3

20.690%Weighted 30%
Source: Epoch AI FrontierMath v2 leaderboardProvenance: Epoch AI reports FrontierMath v2 Tiers 1-3 at 20.69% for claude-opus-4-5-20251101_32K. BenchLM selects the highest published thinking effort for the model and stores the v2 benchmark slice separately.
AIME26Secondary exact

AIME 2026

95.1%Weighted 25%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
HMMT Feb 2026Secondary exact

Harvard-MIT Mathematics Tournament February 2026

85.3%Weighted 25%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
FrontierMath v2 (Tier 4)Benchmark exact

FrontierMath v2 Tier 4

4.167%Weighted 10%
Source: Epoch AI FrontierMath v2 leaderboardProvenance: Epoch AI reports FrontierMath v2 Tier 4 at 4.167% for claude-opus-4-5-20251101_32K. BenchLM selects the highest published thinking effort for the model and stores the v2 benchmark slice separately.
HMMT Feb 2025Secondary exact

Harvard-MIT Mathematics Tournament February 2025

92.9%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
HMMT Nov 2025Secondary exact

Harvard-MIT Mathematics Tournament November 2025

93.3%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
MMAnswerBenchSecondary exact
84.0%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
Multilingual2 benchmarks
MMLU-ProXSecondary exact
85.7%Weighted 100%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
NOVA-63Secondary exact
56.7%Display only
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
Multimodal8 benchmarks
MMMU-ProSecondary exact

Massive Multi-discipline Multimodal Understanding Pro

70.6%Weighted 45%
Source: Qwen3.6-Plus multimodal comparison tableProvenance: Secondary exact
CharXivSecondary exact

CharXiv Reasoning

68.5%Weighted 25%
Source: Qwen3.6-Plus multimodal comparison tableProvenance: Secondary exact
MathVisionSecondary exact
74.3%Display only
Source: Qwen3.6-Plus multimodal comparison tableProvenance: Secondary exact
VideoMMMUSecondary exact
84.4%Display only
Source: Qwen3.6-Plus multimodal comparison tableProvenance: Secondary exact
ScreenSpot ProSecondary exact
45.7%Display only
Source: Qwen3.6-Plus multimodal comparison tableProvenance: Secondary exact
V*Secondary exact
67.0%Display only
Source: Qwen3.6-Plus multimodal comparison tableProvenance: Qwen reports V* in the multimodal comparison table. BenchLM stores the published first value when a slash pair is provided.
AA-MMMU-ProReported

Artificial Analysis MMMU-Pro

71.2%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.
Design Arena WebsiteReported

Design Arena Website Elo

1279Display only
Source: OpenRouter model benchmarksProvenance: Display-only Design Arena Website Elo synced from OpenRouter model benchmark metadata. It is excluded from BenchLM weighted scoring.
Inst. Following3 benchmarks
IFBenchSecondary exact

Instruction Following Benchmark

58%Weighted 65%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
IFEvalSecondary exact

Instruction-Following Eval

90.9%Weighted 35%
Source: Qwen3.6-Plus comparison tableProvenance: Secondary exact
AA-IFBenchReported

Artificial Analysis IFBench

43.0%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.

Claude Opus 4.5 Family

Base entry

Claude Opus 4.5 ThinkingScore 57.44

Frequently Asked Questions

How does Claude Opus 4.5 perform overall in AI benchmarks?

Claude Opus 4.5 currently ranks #34 out of 200 models on BenchLM's provisional leaderboard with an overall score of 64.22. It also ranks #28 out of 99 on the verified leaderboard. It is created by Anthropic. Its published context window is 200K.

Is Claude Opus 4.5 good for knowledge and understanding?

Claude Opus 4.5 ranks #38 out of 52 models in knowledge and understanding benchmarks with an average score of 62.2. There are stronger options in this category.

Is Claude Opus 4.5 good for coding and programming?

Claude Opus 4.5 ranks #20 out of 122 models in coding and programming benchmarks with an average score of 61.9. There are stronger options in this category.

Is Claude Opus 4.5 good for mathematics?

Claude Opus 4.5 ranks #6 out of 7 models in mathematics benchmarks with an average score of 58.6. It is among the top performers in this category.

Is Claude Opus 4.5 good for reasoning and logic?

Claude Opus 4.5 has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.

Is Claude Opus 4.5 good for agentic tool use and computer tasks?

Claude Opus 4.5 ranks #106 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 35.9. There are stronger options in this category.

Is Claude Opus 4.5 good for multimodal and grounded tasks?

Claude Opus 4.5 ranks #26 out of 29 models in multimodal and grounded tasks benchmarks with an average score of 28.4. There are stronger options in this category.

Is Claude Opus 4.5 good for instruction following?

Claude Opus 4.5 ranks #30 out of 38 models in instruction following benchmarks with an average score of 56.4. There are stronger options in this category.

Is Claude Opus 4.5 good for multilingual tasks?

Claude Opus 4.5 ranks #2 out of 13 models in multilingual tasks benchmarks with an average score of 82.9. It is among the top performers in this category.

Which sibling models are related to Claude Opus 4.5?

Claude Opus 4.5 belongs to the Claude Opus 4.5 family. Related variants on BenchLM include Claude Opus 4.5 Thinking.

Does Claude Opus 4.5 have full benchmark coverage on BenchLM?

Not yet. Claude Opus 4.5 currently has 59 published benchmark scores out of the 321 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 Claude Opus 4.5?

Claude Opus 4.5 has a published context window of 200K, which determines how much text it can process in a single interaction.

Related Resources

More from Anthropic LLM Pricing Comparison BenchLM Methodology The AI Race Which LLM Should I Use? Benchmark Explainers
Last updated: July 20, 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.