RankingEffort benchmarks

Claude Opus 4.5

Compare models

65 measured results · 8 effort settings · 46 benchmark/harness combinations

Model evidence and sources →

Change model

Model

Current configurationSources

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare withModel and configuration

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AClaude Opus 4.5 · Thinking 64K3 of 7 capabilities supported

Gaps are unknown, not zero. Hollow points are preliminary.

Agentic
A42.8Preliminary
Hard reasoning
A38.5Supported
Coding
A40.5Supported
Human pref
Unknown
Knowledge
A30.2Supported
Multimodal
Unknown
Long context
Unknown

Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

Capability score evidence

Claude Opus 4.5 · Thinking 64K: 32.9 overall comparison score. Meets the overall evidence requirements.

4 areas measured. The available task mix affects this estimate; missing areas remain unknown.

Which sources carry the weight?

LiveBench · 100.0% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

livebench-coding · 14.3% direct comparison weight
livebench-data · 14.3% direct comparison weight
livebench-language · 14.3% direct comparison weight
multi-swe-bench · 14.3% direct comparison weight
livebench-instructions · 14.3% direct comparison weight
Other families (2) · 28.6% direct comparison weight

These shares describe direct comparison weight, not fractions of the final score or confidence. Source balancing does not give every family equal weight or remove correlations between overlapping sources.

How sensitive is this score?

The audit removes one source at a time and refits against the same overall reference panel. These published results use the matching dataset audited 2026-10-23.

Both ranks use the same surviving configurations for each removal. The cohort starts with 444 configurations supported in the original audit, including separate effort settings. These are diagnostic positions, not the default leaderboard ranks. Eligibility is not rechecked after removal.

What changes when a source is removed?
Removed source	Rank before → after	Rank movement	Cohort remaining / lost support	Score after (change)
Artificial Analysis	99 → 93	Up 6	203 remaining / 241 lost	34.0 (+1.1 points)
ARC Prize	108 → 107	Up 1	444 remaining / 0 lost	36.7 (+3.8 points)
Datacurve	108 → 108	No movement	444 remaining / 0 lost	28.9 (-4.0 points)
LiveCodeBench	108 → 108	No movement	444 remaining / 0 lost	32.9 (+0.0 points)
Scale	108 → 107	Up 1	444 remaining / 0 lost	34.8 (+2.0 points)
Terminal-Bench	108 → 108	No movement	444 remaining / 0 lost	33.0 (+0.1 points)
Cognition	108 → 109	Down 1	444 remaining / 0 lost	33.8 (+0.9 points)
LiveBench	Comparison support lost; no rank		443 remaining / 1 lost	Unavailable
SWE-rebench	108 → 108	No movement	444 remaining / 0 lost	32.0 (-0.9 points)
Epoch AI	106 → 108	Down 2	431 remaining / 13 lost	32.9 (-0.0 points)
Vals AI	108 → 105	Up 3	443 remaining / 1 lost	32.4 (-0.5 points)
Provider reports · Anthropic	108 → 108	No movement	444 remaining / 0 lost	32.9 (+0.0 points)
Provider reports · Moonshot AI	108 → 108	No movement	444 remaining / 0 lost	32.6 (-0.3 points)
Provider reports · Meta	108 → 108	No movement	444 remaining / 0 lost	32.7 (-0.2 points)
Provider reports · Mistral	108 → 108	No movement	444 remaining / 0 lost	32.9 (+0.0 points)
Provider reports · OpenAI	108 → 108	No movement	444 remaining / 0 lost	32.9 (+0.0 points)
All provider reports	108 → 109	Down 1	444 remaining / 0 lost	32.5 (-0.4 points)

Lost support means a configuration no longer has a comparison path to the complete reference panel. Its rank is unavailable. A surviving path alone does not establish leaderboard eligibility.

Overlapping sources may remain correlated even with source balancing. Leave-one-source-out movement is a sensitivity diagnostic, not a spread of source-specific ranks or a confidence measure. Scoring methodology · Dated configuration audit

Every benchmark family

Removed evidence	Recomputed score	Change
multi-swe-bench	37.0	+4.1
deepswe	28.9	-4.0
livebench-coding	28.9	-4.0
livebench-language	30.3	-2.6
livebench-instructions	35.3	+2.4
arc-agi	35.0	+2.1
livebench-reasoning	34.9	+2.0
livebench-math	31.1	-1.8
hle	34.2	+1.3
mmmu-pro	34.1	+1.2
frontiercode	33.8	+0.9
swe-rebench	32.0	-0.9
arc-agi-3	33.5	+0.6
mmlu-pro	33.3	+0.4
terminal-bench	33.3	+0.4
epoch-game-puzzles	32.6	-0.3
harvey	32.6	-0.3
aime	32.7	-0.2
swe-atlas-qna	32.7	-0.2
omniscience	33.1	+0.2
enigma-eval	32.7	-0.2
ifbench	32.7	-0.2
vals-finance-agent	32.7	-0.2
vibe-code	32.7	-0.2
swe-atlas-test-writing	32.7	-0.2
tau-bench	32.8	-0.1
briefcase	32.8	-0.1
gdp-pdf	33.0	+0.1
simpleqa	32.8	-0.1
aa-lcr	33.0	+0.1
swe-atlas-refactoring	32.8	-0.1
gdpval	32.8	-0.0
vals-excel-modeling	32.8	-0.0
analyst-agent	32.8	-0.0
terminal-bench-science	32.8	-0.0
automationbench	32.9	-0.0
vals-code-migration	32.9	+0.0
apex-agents	32.9	+0.0
frontiermath	32.9	-0.0
critpt	32.9	+0.0
swe-bench-pro	32.9	+0.0
vals-tax-agent	32.9	+0.0
gpqa	32.9	-0.0
enterprise-ops	32.9	+0.0
itbench	32.9	-0.0
livecodebench	32.9	+0.0
vals-legal-research	32.9	+0.0
gmmlu	32.9	+0.0
livebench-data	32.9	-0.0
math500	32.9	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 68 observations, 55 contributing, 0 matched without graph weight and 13 excluded. Counts do not establish rank eligibility.

3 · Benchmark family or source protocol has not been reviewed
3 · Effort is not identified
4 · Fewer than two distinct model identities remain in this evaluation
3 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately

Inspect every observation and exclusion →

Benchmark results

These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

Benchmark / source	Thinking 64K	Thinking	Thinking 16K	Thinking 32K	Thinking 48K	Reasoning	Non-reasoning	Unspecified
AIME 2025Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	91.33% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	62.67% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
AA-LCR v1.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	77.33% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	70.67% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
CritPtArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Argonne and UIUC, with contributions from 60+ researchers globally. CritPt is under review by its authors following external feedback · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	4.57% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	0.29% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
Global-MMLU-LiteArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	91.28% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	90.59% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
GPQA DiamondArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	86.57% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	81.01% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	30.12% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	13.16% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
IFBenchArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	57.96% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	42.99% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
LiveCodeBenchArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	87.09% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	73.76% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
MMLU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	89.45% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	88.92% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
MMMU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	74.05% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	71.21% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
AA-Omniscience IndexArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis Unit: index. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	14 Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	-4.13 Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
AA-Omniscience AccuracyArtificial Analysis · 6000 questions; 1 attempt; correct-answer share; GPT-5.6 Luna medium grader Benchmark details AA-Omniscience Accuracy (higher is better) measures the proportion of correctly answered questions out of all questions, regardless of whether the model chooses to answer · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	46.58% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	40.92% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
𝜏²-Bench TelecomArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	89.47% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	86.26% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
Terminal-Bench HardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	Not reported	Not reported	46.97% Reported settings Claude Opus 4.5 (Reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5-thinking	40.91% Reported settings Claude Opus 4.5 (Non-reasoning); Model source: https://artificialanalysis.ai/models/claude-opus-4-5	Not reported
HLE · text only · Scale official evaluationScale · Scale official evaluation Benchmark details Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging. Unit: percent. Source collected 2026-10-18.	Not reported	26.32% Reported settings Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging.; published row 2025-11-26T20:27:42.000Z; reported margin ±1.86 percentage points; Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
HLE · full · Scale official evaluationScale · Scale official evaluation Benchmark details Scale full public HLE; 2,500 questions; automatic answer judging; temperature 0 when configurable. Public-set contamination caveats retained per result. Unit: percent. Source collected 2026-10-18.	Not reported	25.2% Reported settings Scale full public HLE; 2,500 questions; automatic answer judging; temperature 0 when configurable. Public-set contamination caveats retained per result.; published row 2025-11-26T20:23:24.000Z; reported margin ±1.7 percentage points; Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · AMPS Hard · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	99% Reported settings LiveBench official task evaluation; livebench-2026-06-25-AMPS_Hard; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · code completion · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	80.44% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_completion; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · code generation · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	78.87% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_generation; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · connections · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	99.33% Reported settings LiveBench official task evaluation; livebench-2026-06-25-connections; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · consecutive events · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	79.36% Reported settings LiveBench official task evaluation; livebench-2026-06-25-consecutive_events; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · integrals with game · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	78% Reported settings LiveBench official task evaluation; livebench-2026-06-25-integrals_with_game; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · javascript · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	59.09% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-javascript; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · logic with navigation · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	68% Reported settings LiveBench official task evaluation; livebench-2026-06-25-logic_with_navigation; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
LiveBench · math comp · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-10-18.	95.1% Reported settings LiveBench official task evaluation; livebench-2026-06-25-math_comp; claude-opus-4-5-20251101-thinking-64k-high-effort; checkpoint=claude-opus-4-5-20251101	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported

What these comparisons mean

Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

Download all effort benchmark results · 8,916 results across 259 models · Collected 2026-10-18

Claude Opus 4.5

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results