Fronset Benchmark

Qwen 3.6 Plus (archived)

Provider
Alibaba Cloud (DashScope)
Model name
qwen3.6-plus
Last benchmarked
Qualified on
24 / 51 tasks (at the 90% bar)
Best value on
1 tasks

Good enough on 24/51 tasks at the 90% bar. Best value on 1 task. Best fit for: Structured Data & Fact Extraction.

Cost vs quality across all tasks, as of 15 August 2026

0%25%50%75%100%78910Quality score (7–10)Cost-efficiency vs best value (1.0 = best value)Engagement Triage — quality 8.50, 36.2x the cost of the best-value good-enough optionAuthor Matching — quality 7.81, 16.2x the cost of the best-value good-enough optionClaim Extraction — quality 7.04, 14.4x the cost of the best-value good-enough optionOnboarding Chapter Prompt Adaptation — quality 8.66, 18.7x the cost of the best-value good-enough optionImage Prompt Generation — quality 8.58, 18.0x the cost of the best-value good-enough optionVetted News Site Selection — quality 8.11, this model IS the best-value good-enough option★X Post Selection — quality 8.52, 8.2x the cost of the best-value good-enough optionInvestment Panel Voting — quality 8.60, 10.5x the cost of the best-value good-enough optionSEC S-1 Chunk Analysis — quality 8.67, 20.7x the cost of the best-value good-enough optionTopic Grouping and Client Matching — quality 7.85, 9.6x the cost of the best-value good-enough optionPublication Title Generation — quality 8.29, 16.8x the cost of the best-value good-enough optionAuthor Living-Person Safety Check — quality 8.62, 13.1x the cost of the best-value good-enough optionLanguage Detection — quality 9.91, 57.5x the cost of the best-value good-enough optionResearch Query Generation — quality 7.72, 23.7x the cost of the best-value good-enough optionSubstack Newsletter — quality 8.96, 19.6x the cost of the best-value good-enough optionS-1 TOC Extraction — quality 8.79, 11.3x the cost of the best-value good-enough optionStructured Output Extraction — quality 9.65, 58.0x the cost of the best-value good-enough optionClaim Refinement — quality 8.19, 14.4x the cost of the best-value good-enough optionTranslation — quality 8.48, 38.7x the cost of the best-value good-enough optionPrompt Adaptation — quality 8.81, 34.3x the cost of the best-value good-enough optionTopic Cluster Naming — quality 8.23, 14.7x the cost of the best-value good-enough optionClaim-Referenced Analyst Writing — quality 8.30, 8.8x the cost of the best-value good-enough optionActivity Feed Blurb Generation — quality 8.34, 46.2x the cost of the best-value good-enough optionTopic-to-Section Assignment — quality 8.17, 9.8x 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 15 August 2026

TaskCategoryQuality (% of best)ConfidenceOverpay
Vetted News Site Selection ★Relevance, Classification & Matching94%HIGHbest value
X Post SelectionRelevance, Classification & Matching96%RANKED8.2x
Claim-Referenced Analyst WritingLong-form Content Generation90%HIGH8.8x
Topic Grouping and Client MatchingRelevance, Classification & Matching92%HIGH9.6x
Topic-to-Section AssignmentTopic Organization & Clustering93%MEDIUM9.8x
Investment Panel VotingFinancial Analysis & Trading Decisions95%RANKED10x
S-1 TOC ExtractionStructured Data & Fact Extraction94%MEDIUM11x
Author Living-Person Safety CheckRelevance, Classification & Matching93%MEDIUM13x
Claim ExtractionStructured Data & Fact Extraction91%HIGH14x
Claim RefinementInfrastructure & Utility97%HIGH14x
Topic Cluster NamingTopic Organization & Clustering92%HIGH15x
Author MatchingRelevance, Classification & Matching92%MEDIUM16x
Publication Title GenerationContent Summarization & Synthesis91%RANKED17x
Image Prompt GenerationInfrastructure & Utility92%RANKED18x
Onboarding Chapter Prompt AdaptationInfrastructure & Utility96%RANKED19x
Substack NewsletterLong-form Content Generation95%HIGH20x
SEC S-1 Chunk AnalysisFinancial Analysis & Trading Decisions94%RANKED21x
Research Query GenerationInfrastructure & Utility90%MEDIUM24x
Prompt AdaptationInfrastructure & Utility99%RANKED34x
Engagement TriageRelevance, Classification & Matching92%HIGH36x
TranslationInfrastructure & Utility90%HIGH39x
Activity Feed Blurb GenerationSocial & Promotional Content95%MEDIUM46x
Language DetectionRelevance, Classification & Matching99%RANKED57x
Structured Output ExtractionStructured Data & Fact Extraction98%RANKED58x

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.