The Right-Sized LLM Benchmark Cheapest LLM that's good enough.

Meta Muse Spark 1.1 at a glance

Good enough on 27/52 tasks at the 90% bar. Best value on 2 tasks. Best fit for: Social & Promotional Content, Relevance, Classification & Matching, Topic Organization & Clustering. Doesn't qualify on

Provider
Meta
Model name
muse-spark-1.1
Cost mode:
Qualifies on
27 / 52 tasks (at 90% bar)
Best value on
2 tasks

Cost vs quality across all tasks

0%25%50%75%100%678910Quality score (6–10)Cost-efficiency vs best value (1.0 = best value)Vetted News Site Selection — quality 7.72, 5.7x the cost of the best-value good-enough optionOnboarding Subject Analysis — quality 9.22, 9.2x the cost of the best-value good-enough optionContent Domain Suggestion — quality 8.80, this model IS the best-value good-enough option★Topic Sequence Ordering — quality 8.78, 7.6x the cost of the best-value good-enough optionX Post Relevance Scoring — quality 6.97, 5.6x the cost of the best-value good-enough optionActivity Feed Blurb Generation — quality 8.47, 36.6x the cost of the best-value good-enough optionSubstack Newsletter — quality 8.82, 45.7x the cost of the best-value good-enough optionInvestment Panel Voting — quality 8.76, 3.7x the cost of the best-value good-enough optionSocial Post Promotion — quality 8.76, 23.6x the cost of the best-value good-enough optionAuthor Living-Person Safety Check — quality 9.08, 25.3x the cost of the best-value good-enough optionTopic Discovery Clustering — quality 9.06, this model IS the best-value good-enough option★Language Detection — quality 9.84, 75.7x the cost of the best-value good-enough optionExecutive Summary Generation — quality 7.55, 10.9x the cost of the best-value good-enough optionTopic Report Relevance Scoring — quality 8.95, 5.1x the cost of the best-value good-enough optionPublication Title Generation — quality 8.75, 27.0x the cost of the best-value good-enough optionLanding Page Section Generation — quality 9.14, 9.3x the cost of the best-value good-enough optionTopic Cluster Naming — quality 8.36, 9.1x the cost of the best-value good-enough optionTopic Grouping and Client Matching — quality 8.28, 14.1x the cost of the best-value good-enough optionEngagement Triage — quality 8.81, 56.9x the cost of the best-value good-enough optionPrompt Adaptation — quality 8.68, 26.7x the cost of the best-value good-enough optionOnboarding Chapter Outline Generation — quality 8.88, 11.7x the cost of the best-value good-enough optionX Post Selection — quality 8.38, 96.8x the cost of the best-value good-enough optionAuthor Voice Generation — quality 9.18, 15.3x the cost of the best-value good-enough optionSEC S-1 Chunk Analysis — quality 9.20, 10.0x the cost of the best-value good-enough optionImage Prompt Generation — quality 8.89, 29.7x the cost of the best-value good-enough optionEngagement Reply Draft — quality 7.79, 11.2x the cost of the best-value good-enough optionTopic-to-Section Assignment — quality 8.33, 36.2x 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

TaskCategoryQuality (% of best)ConfidenceOverpay
Topic Discovery Clustering ★Topic Organization & Clustering97%MEDIUMbest value
Content Domain Suggestion ★ bestRelevance, Classification & Matching100%MEDIUMbest value
Investment Panel VotingFinancial Analysis & Trading Decisions96%MEDIUM3.7x
Topic Report Relevance ScoringRelevance, Classification & Matching96%MEDIUM5.1x
X Post Relevance ScoringRelevance, Classification & Matching94%MEDIUM5.6x
Vetted News Site SelectionRelevance, Classification & Matching91%HIGH5.7x
Topic Sequence OrderingTopic Organization & Clustering98%HIGH7.6x
Topic Cluster NamingTopic Organization & Clustering96%MEDIUM9.1x
Onboarding Subject Analysis bestFinancial Analysis & Trading Decisions100%HIGH9.2x
Landing Page Section GenerationLong-form Content Generation93%MEDIUM9.3x
SEC S-1 Chunk AnalysisFinancial Analysis & Trading Decisions100%HIGH10x
Executive Summary GenerationContent Summarization & Synthesis95%MEDIUM11x
Engagement Reply DraftSocial & Promotional Content92%MEDIUM11x
Onboarding Chapter Outline GenerationLong-form Content Generation91%HIGH12x
Topic Grouping and Client MatchingRelevance, Classification & Matching96%HIGH14x
Author Voice GenerationLong-form Content Generation95%RANKED15x
Social Post PromotionSocial & Promotional Content99%RANKED24x
Author Living-Person Safety CheckRelevance, Classification & Matching98%HIGH25x
Prompt AdaptationInfrastructure & Utility98%HIGH27x
Publication Title GenerationContent Summarization & Synthesis97%HIGH27x
Image Prompt GenerationInfrastructure & Utility97%HIGH30x
Topic-to-Section AssignmentTopic Organization & Clustering93%MEDIUM36x
Activity Feed Blurb GenerationSocial & Promotional Content96%RANKED37x
Substack NewsletterLong-form Content Generation95%RANKED46x
Engagement TriageRelevance, Classification & Matching99%HIGH57x
Language DetectionRelevance, Classification & Matching98%HIGH76x
X Post SelectionRelevance, Classification & Matching96%HIGH97x

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.