Fronset Benchmark

Claude Opus 5 (archived)

Provider
Anthropic
Model name
claude-opus-5
Last benchmarked
Qualified on
17 / 53 tasks (at the 90% bar)
Best value on
1 tasks

Good enough on 17/53 tasks at the 90% bar. Best value on 1 task. Best fit for: Infrastructure & Utility, Relevance, Classification & Matching, Social & Promotional Content, Content Summarization & Synthesis.

Cost vs quality across all tasks, as of 24 September 2026

0%25%50%75%100%78910Quality score (7–10)Cost-efficiency vs best value (1.0 = best value)Publication Title Package Generation — quality 8.59, 3.4x the cost of the best-value good-enough optionTopic Cluster Labeling — quality 8.91, 9.4x the cost of the best-value good-enough optionContent Domain Suggestion — quality 9.18, this model IS the best-value good-enough option★Bench Hypothesis Post Generation — quality 9.19, 23.7x the cost of the best-value good-enough optionAtomic Fact Claim Extraction — quality 7.98, 10.4x the cost of the best-value good-enough optionSocial Post Portfolio Selection — quality 8.64, 31.8x the cost of the best-value good-enough optionProfile Pool Matching — quality 8.50, 3.9x the cost of the best-value good-enough optionMulti Perspective Decision Synthesis — quality 8.85, 26.8x the cost of the best-value good-enough optionPersona Policy Eligibility Check — quality 9.04, 10.1x the cost of the best-value good-enough optionPromotional Message Generation — quality 8.65, 12.0x the cost of the best-value good-enough optionBatch Text Translation — quality 9.05, 35.2x the cost of the best-value good-enough optionAuthoritative Source Selection — quality 8.34, 32.8x the cost of the best-value good-enough optionFactual Claim Refinement — quality 8.18, 19.9x the cost of the best-value good-enough optionLanguage Identification — quality 9.80, 27.7x the cost of the best-value good-enough optionEngagement Opportunity Triage — quality 8.61, 26.8x the cost of the best-value good-enough optionResearch Query Generation — quality 7.35, 51.3x the cost of the best-value good-enough optionStructured Content Summarization — quality 8.91, 21.8x the cost of the best-value good-enough option

within ~1.3× of the best-value model · 1.3–2× · >2× · ★ this model is the best-value pick on that task. Top-right = best quadrant. Only tasks where this model qualifies at the 90% bar are plotted.

Per-task breakdown, as of 24 September 2026

TaskCategoryQuality (% of best)ConfidenceOverpay
Content Domain Suggestion ★ bestRelevance, Classification & Matching100%MEDIUMbest value
Publication Title Package GenerationContent Summarization & Synthesis94%HIGH3.4x
Profile Pool MatchingRelevance, Classification & Matching96%MEDIUM3.9x
Topic Cluster Labeling bestTopic Organization & Clustering100%MEDIUM9.4x
Persona Policy Eligibility CheckRelevance, Classification & Matching98%MEDIUM10x
Atomic Fact Claim Extraction bestStructured Data & Fact Extraction100%MEDIUM10x
Promotional Message GenerationSocial & Promotional Content99%MEDIUM12x
Factual Claim RefinementInfrastructure & Utility95%MEDIUM20x
Structured Content SummarizationContent Summarization & Synthesis96%HIGH22x
Bench Hypothesis Post Generation bestSocial & Promotional Content100%RANKED24x
Engagement Opportunity TriageRelevance, Classification & Matching97%RANKED27x
Multi Perspective Decision Synthesis bestFinancial Analysis & Trading Decisions100%RANKED27x
Language IdentificationRelevance, Classification & Matching98%MEDIUM28x
Social Post Portfolio SelectionRelevance, Classification & Matching98%HIGH32x
Authoritative Source SelectionRelevance, Classification & Matching95%MEDIUM33x
Batch Text TranslationInfrastructure & Utility96%RANKED35x
Research Query GenerationInfrastructure & Utility90%MEDIUM51x

Overpay — how much more you pay by running this model instead of the best-value model that clears the quality bar on that task (marked ★). "16x" means you overpay 16× — the same output for 16× the best-value good-enough option; ★ means this model is that option (no overpayment). Confidence — how sure we are about the quality score (more judgments + more agreement = higher confidence): RANKED many independent judges scored this model's outputs and their agreement is very high (most confident) — HIGH many judges have scored it and they mostly agree (well-pinned) — MEDIUM enough judges have weighed in to publish, but they disagree more than we'd like (treat with a small grain of salt). LOW-confidence cells are hidden everywhere on the site. See the methodology for the exact thresholds.