New: Agent Index for knowledge work is live.Explore the index ↗
Compare models

Across the task benchmarks we track, Claude Opus 4.8’s strongest relative results are in reasoning, knowledge work, and work automations. On GPQA Diamond, it scores 92%, ranking #13 among 30 models with reported results in our dataset.

Updated: Data as of:
Compare this model

Strongest results

Strongest results for Claude Opus 4.8 in tracked evaluation cohorts
BenchmarkMeasuresScore
Humanity's Last Exam ↗Stable HLE — text-only subsetExpert reasoning48.7%Joint #10 of 68
GDP.pdf ↗100 tasks / All-pass / five attemptsProfessional document reasoning22.8%Joint #15 of 55
GPQA ↗Diamond / Artificial AnalysisReasoning92.02%#13 of 30
AA-Briefcase ↗1.1 / combined EloBusiness deliverables1,320 Elo#30 of 57

Model details

Proprietary11 benchmarks
Specifications for Claude Opus 4.8
ModelClaude Opus 4.8
Agent Index v1.2Insufficient coverage (3/7) · Preview
CreatorAnthropic (opens in a new tab)
AvailabilityProprietary
Context size1,000,000 tokens (opens in a new tab)
Max output128,000 tokens (opens in a new tab)
Knowledge cutoffNot reported
LicenseNot reported

Pricing and performance

Per provider and reported configuration. Prices are USD per 1M tokens.

Provider offerings for Claude Opus 4.8
Provider / configurationContexttokensInputUSD / 1M tokensOutputUSD / 1M tokensLatencyTTFT · secondsSpeedtokens / sec
Amazon (opens in a new tab)
Claude Opus 4.8 (max)		1,000,000$5.00 (opens in a new tab)$25.00 (opens in a new tab)49.238 (opens in a new tab)70.462 (opens in a new tab)
Anthropic (opens in a new tab)
Claude Opus 4.8 (max)		1,000,000$5.00 (opens in a new tab)$25.00 (opens in a new tab)56.523 (opens in a new tab)65.881 (opens in a new tab)
Google (opens in a new tab)
Claude Opus 4.8 (max)		1,000,000$5.00 (opens in a new tab)$25.00 (opens in a new tab)57.41 (opens in a new tab)67.892 (opens in a new tab)

Benchmark performance

Claude Opus 4.8 in color, alongside leading models. Latest stored reports, best effort per report date.

Reported benchmarks

All published results in our dataset, with versions, evaluation settings, and sources.

All published benchmark reports for Claude Opus 4.8
BenchmarkScoreEvaluation setupSource
AA-Briefcase1.1 / combined EloCombined Elo · Higher is better1,320 Eloselection: best reported configuration per release · Claude Opus 4.8 (max) · tools: benchmark-definedArtificial Analysis (opens in a new tab)independent
AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better13.55%Claude Opus 4.8 (Low); private held-out 657 tasks; API mode; 50-step cap; strict completionZapier ↗ (opens in a new tab)benchmark publisher
AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better14%Claude Opus 4.8 (Medium); private held-out 657 tasks; API mode; 50-step cap; strict completionZapier ↗ (opens in a new tab)benchmark publisher
AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better16.13%Claude Opus 4.8 (High); private held-out 657 tasks; API mode; 50-step cap; strict completionZapier ↗ (opens in a new tab)benchmark publisher
AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better16.44%Claude Opus 4.8 (Max); private held-out 657 tasks; API mode; 50-step cap; strict completionZapier ↗ (opens in a new tab)benchmark publisher
AutomationBench1.0.6 / Zapier held-out private setStrict task completion · Higher is better16.89%Claude Opus 4.8 (XHigh); private held-out 657 tasks; API mode; 50-step cap; strict completionZapier ↗ (opens in a new tab)benchmark publisher
AutomationBench-AA1.0.6 / AA private 657Objectives completed without guardrail violations · Higher is better45.6%reported precision decimals: 1 · result quality: reported · Claude Opus 4.8 (max); private held-out 657 tasks; 50 turns; objective completion with zero credit for any guardrail violation. · source model href: /models/claude-opus-4-8 · task count: 657 · Tools enabled · turn cap: 50Artificial Analysis (opens in a new tab)independent
DeepSWEv1.1Pass@1 · Higher is better59%mini-swe-agent; effort=max; ±2 percentage pointsDatacurve ↗ (opens in a new tab)benchmark publisher
GDP.pdf100 tasks / All-pass / five attemptsAccuracy · Higher is better22.8%selection: best reported configuration per release · Claude Opus 4.8 (max) · tools: benchmark-definedArtificial Analysis (opens in a new tab)independent
GDPval-AA2.1Elo · Higher is better1,461.75 EloClaude Opus 4.8 (Max); 220 tasks; Stirrup agent; pairwise EloArtificial Analysis (opens in a new tab)independent
GPQADiamond / Artificial AnalysisAccuracy · Higher is better92.02%Claude Opus 4.8 (Max); 198-question Diamond subset; independent AA evaluationArtificial Analysis (opens in a new tab)independent
Humanity's Last ExamStable HLE — full setAccuracy · Higher is better49.8%subset: full · No toolsZ.ai ↗ (opens in a new tab)other · 2026-06-16
Humanity's Last ExamStable HLE — full setAccuracy · Higher is better57.9%subset: full · Tools enabledZ.ai ↗ (opens in a new tab)other · 2026-06-16
Humanity's Last ExamStable HLE — text-only subsetAccuracy · Higher is better48.7%question count: 2158 · reported precision decimals: 1 · Claude Opus 4.8 (max); original HLE; 2,158 text-only questions; no tools; pass@1; chart labels rounded to one decimal. · source model href: /models/claude-opus-4-8 · subset: text · No toolsArtificial Analysis (opens in a new tab)independent
Terminal-Bench4.0Task success · Higher is better23.6%Claude Code; effort=max; 23.6% ± 3.6%Terminal-Bench ↗ (opens in a new tab)benchmark publisher
Terminal-Bench4.0 / mini-swe-agentTask success · Higher is better21.717%harness: mini-swe-agent · repeats: 3 · selection: best reported configuration per release · Claude Opus 4.8 (Max)Artificial Analysis (opens in a new tab)independent