Fronset Benchmark

Claude Sonnet 5 (archived)

Provider
Anthropic
Model name
claude-sonnet-5
Last benchmarked
Qualified on
25 / 53 tasks (at the 90% bar)
Best value on
3 tasks

Good enough on 25/53 tasks at the 90% bar. Best value on 3 tasks. Best fit for: Financial Analysis & Trading Decisions, Infrastructure & Utility, Long-form Content Generation, Relevance, Classification & Matching.

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)Reference Preserving Analytical Writing — quality 8.88, 14.3x the cost of the best-value good-enough optionGenerated Content Relevance Scoring — quality 8.44, 3.9x the cost of the best-value good-enough optionPublication Title Package Generation — quality 8.75, 1.9x the cost of the best-value good-enough optionReport Executive Summary Generation — quality 8.21, 2.4x the cost of the best-value good-enough optionProfiled Document Analysis — quality 8.21, 1.6x the cost of the best-value good-enough optionVoice Profile Generation — quality 8.92, 9.5x the cost of the best-value good-enough optionPublic Engagement Reply Review — quality 8.61, this model IS the best-value good-enough option★Subject Configuration Generation — quality 8.89, 8.1x the cost of the best-value good-enough optionProfiled Document Section Analysis — quality 8.76, 28.2x the cost of the best-value good-enough optionTopic Cluster Labeling — quality 8.37, 8.0x the cost of the best-value good-enough optionCommunity Policy Evaluation — quality 9.27, 11.7x the cost of the best-value good-enough optionBench Hypothesis Post Generation — quality 8.48, 10.8x the cost of the best-value good-enough optionAtomic Fact Claim Extraction — quality 7.66, 4.0x the cost of the best-value good-enough optionResearch Region Identification — quality 8.77, this model IS the best-value good-enough option★Content To Image Prompt Generation — quality 8.63, 12.4x the cost of the best-value good-enough optionSocial Post Portfolio Selection — quality 8.33, 16.4x the cost of the best-value good-enough optionCommunity Content Promotion Generation — quality 8.19, 20.4x the cost of the best-value good-enough optionVisual Theme Configuration Generation — quality 9.41, 25.5x the cost of the best-value good-enough optionMetadata Paragraph Rewriting — quality 9.42, this model IS the best-value good-enough option★Newsletter Copy Generation — quality 9.32, 6.0x the cost of the best-value good-enough optionMulti Perspective Decision Synthesis — quality 8.30, 15.3x the cost of the best-value good-enough optionBatch Text Translation — quality 8.95, 12.3x the cost of the best-value good-enough optionFactual Claim Refinement — quality 8.35, 8.2x the cost of the best-value good-enough optionLanguage Identification — quality 10.03, 20.9x the cost of the best-value good-enough optionEngagement Opportunity Triage — quality 8.55, 16.1x 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
Public Engagement Reply Review ★Relevance, Classification & Matching91%MEDIUMbest value
Research Region Identification ★Structured Data & Fact Extraction99%RANKEDbest value
Metadata Paragraph Rewriting ★ bestInfrastructure & Utility100%RANKEDbest value
Profiled Document AnalysisFinancial Analysis & Trading Decisions94%HIGH1.6x
Publication Title Package GenerationContent Summarization & Synthesis95%RANKED1.9x
Report Executive Summary GenerationContent Summarization & Synthesis94%RANKED2.4x
Generated Content Relevance ScoringRelevance, Classification & Matching93%MEDIUM3.9x
Atomic Fact Claim ExtractionStructured Data & Fact Extraction96%MEDIUM4x
Newsletter Copy Generation bestLong-form Content Generation100%RANKED6x
Topic Cluster LabelingTopic Organization & Clustering94%RANKED8x
Subject Configuration Generation bestFinancial Analysis & Trading Decisions100%HIGH8.1x
Factual Claim RefinementInfrastructure & Utility97%HIGH8.2x
Voice Profile GenerationLong-form Content Generation91%RANKED9.5x
Bench Hypothesis Post GenerationSocial & Promotional Content92%MEDIUM11x
Community Policy Evaluation bestRelevance, Classification & Matching100%MEDIUM12x
Batch Text TranslationInfrastructure & Utility95%RANKED12x
Content To Image Prompt GenerationInfrastructure & Utility96%RANKED12x
Reference Preserving Analytical WritingLong-form Content Generation97%RANKED14x
Multi Perspective Decision SynthesisFinancial Analysis & Trading Decisions94%MEDIUM15x
Engagement Opportunity TriageRelevance, Classification & Matching96%RANKED16x
Social Post Portfolio SelectionRelevance, Classification & Matching95%RANKED16x
Community Content Promotion GenerationSocial & Promotional Content95%RANKED20x
Language IdentificationRelevance, Classification & Matching100%RANKED21x
Visual Theme Configuration Generation bestLong-form Content Generation100%RANKED26x
Profiled Document Section AnalysisFinancial Analysis & Trading Decisions96%MEDIUM28x

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.