Cost mode:

At a glance

Good enough on 3/35 tasks at the 95% bar. Cheapest qualifier on 0 tasks. Doesn't qualify on any: Financial Analysis & Trading Decisions, Structured Data & Fact Extraction, Content Summarization & Synthesis, Social & Promotional Content, Infrastructure & Utility.

Provider
Anthropic
Model name
claude-haiku-4-5
Qualifies on
3 / 35 tasks (at 90% bar)
Cheapest qualifier on
0 tasks

Cost vs quality across all tasks

$0.00021$0.00102$0.00489$0.02343$0.112290246810Quality score (0–10)Blended cost / callat_content_domain_suggest autogenerated — quality 0.00, cost $0.002906Author Matching — quality 6.62, cost $0.007162author_soul_generation autogenerated — quality 9.14, cost $0.023960auto_reddit_post_generation autogenerated — quality 7.49, cost $0.061540claim_extraction autogenerated — quality 6.63, cost $0.002858claim_refinement autogenerated — quality 6.31, cost $0.002920Claim-Referenced Analyst Writing (pooled) — quality 8.02, cost $0.017526content-summarization autogenerated — quality 6.47, cost $0.008622executive_summary_generation autogenerated — quality 0.00, cost $0.035708Generic TOC Extraction — quality 0.00, cost $0.001750image_prompt_generation autogenerated — quality 7.59, cost $0.005360Language Detection — quality 9.99, cost $0.000213LLM Prompt Adaptation — quality 7.38, cost $0.014782markdown_newline_repair autogenerated — quality 1.50, cost $0.112286metadata_paragraph_improvement autogenerated — quality 5.17, cost $0.001074onboarding_chapter_generation autogenerated — quality 0.00, cost $0.032188onboarding_chapter_prompt_generation autogenerated — quality 8.22, cost $0.072700onboarding_prospect_analysis autogenerated — quality 0.00, cost $0.004592ps_section_reassignment autogenerated — quality 0.00, cost $0.008722region_identification autogenerated — quality 7.92, cost $0.004180Relevance Scoring (POST) — quality 3.28, cost $0.001750Relevance Scoring (Topic Report) — quality 7.19, cost $0.001750Relevance Scoring (X Post) — quality 6.09, cost $0.001750S1 TOC extraction — quality 0.00, cost $0.006824SEC Filling Analysis — quality 5.98, cost $0.012259sec-s1-chunk-analysis autogenerated — quality 7.67, cost $0.020164section_generation autogenerated — quality 6.73, cost $0.006250Social Post Promo (pooled) — quality 7.39, cost $0.003608structured_output_extraction autogenerated — quality 8.38, cost $0.008306subreddit_selection autogenerated — quality 4.60, cost $0.003952subreddit_vetting autogenerated — quality 6.50, cost $0.006025Substack Newsletter (pooled) — quality 8.87, cost $0.011004synthesis_analysis autogenerated — quality 7.64, cost $0.021765synthesis_of_titles_for_publication autogenerated — quality 8.01, cost $0.006009theme_generation autogenerated — quality 0.00, cost $0.010548Topic Discovery Clustering (pooled) — quality 7.13, cost $0.061034topic_client_matching autogenerated — quality 5.09, cost $0.019231topic_cluster_naming autogenerated — quality 8.22, cost $0.006792topic_sequence_determination autogenerated — quality 0.00, cost $0.006870Trading Recommendation — quality 8.24, cost $0.046094Translation — quality 6.34, cost $0.005402vetted_site_selection autogenerated — quality 6.88, cost $0.023179x_com_messages_for_promotion autogenerated — quality 0.00, cost $0.032237x_post_selection autogenerated — quality 7.79, cost $0.003228

qualifies at 90% bar · doesn't qualify · ★ this model is the best on that task. Lower + further right = cheaper + higher quality. Y-axis is log-scaled.

Per-task breakdown

CategoryTaskQualityConfidenceCost / callvs bestQualifies @ 90%
Structured Data & Fact Extractionclaim_extraction autogenerated6.63high · n=100$0.002858-422%no
Structured Data & Fact ExtractionGeneric TOC Extraction0.00low · n=0$0.001750no
Structured Data & Fact Extractionregion_identification autogenerated7.92low · n=25$0.00418039%no
Structured Data & Fact ExtractionS1 TOC extraction0.00low · n=0$0.006824no
Structured Data & Fact Extractionstructured_output_extraction autogenerated8.38medium · n=91$0.008306-171%no
Financial Analysis & Trading Decisionsonboarding_prospect_analysis autogenerated0.00low · n=0$0.004592no
Financial Analysis & Trading DecisionsSEC Filling Analysis5.98medium · n=100$0.01225983%no
Financial Analysis & Trading Decisionssec-s1-chunk-analysis autogenerated7.67high · n=81$0.02016482%no
Financial Analysis & Trading Decisionssynthesis_analysis autogenerated7.64medium · n=29$0.02176567%no
Financial Analysis & Trading DecisionsTrading Recommendation8.24high · n=86$0.04609467%no
Infrastructure & Utilityclaim_refinement autogenerated6.31medium · n=100$0.002920-40%no
Infrastructure & Utilityimage_prompt_generation autogenerated7.59ranked · n=100$0.00536080%no
Infrastructure & UtilityLLM Prompt Adaptation7.38high · n=100$0.01478267%no
Infrastructure & Utilitymarkdown_newline_repair autogenerated1.50low · n=1$0.112286no
Infrastructure & Utilitymetadata_paragraph_improvement autogenerated5.17medium · n=90$0.001074-32%no
Infrastructure & Utilityonboarding_chapter_prompt_generation autogenerated8.22high · n=75$0.07270083%no
Infrastructure & UtilityTranslation6.34medium · n=93$0.00540283%no
Long-form Content Generationauthor_soul_generation autogenerated9.14ranked · n=84$0.02396080%
Long-form Content GenerationClaim-Referenced Analyst Writing (pooled)8.02high · n=37$0.01752667%no
Long-form Content Generationonboarding_chapter_generation autogenerated0.00low · n=0$0.032188no
Long-form Content Generationsection_generation autogenerated6.73high · n=6$0.006250-29%no
Long-form Content GenerationSubstack Newsletter (pooled)8.87ranked · n=5$0.01100480%no
Long-form Content Generationtheme_generation autogenerated0.00low · n=0$0.010548no
Relevance, Classification & Matchingat_content_domain_suggest autogenerated0.00low · n=0$0.002906no
Relevance, Classification & MatchingAuthor Matching6.62high · n=100$0.00716271%no
Relevance, Classification & MatchingLanguage Detection9.99ranked · n=82$0.000213-90%
Relevance, Classification & MatchingRelevance Scoring (POST)3.28high · n=100$0.00175056%no
Relevance, Classification & MatchingRelevance Scoring (Topic Report)7.19medium · n=44$0.001750-994%no
Relevance, Classification & MatchingRelevance Scoring (X Post)6.09low · n=11$0.001750no
Relevance, Classification & Matchingsubreddit_selection autogenerated4.60high · n=63$0.00395281%no
Relevance, Classification & Matchingsubreddit_vetting autogenerated6.50low · n=1$0.006025no
Relevance, Classification & Matchingtopic_client_matching autogenerated5.09medium · n=100$0.019231-523%no
Relevance, Classification & Matchingvetted_site_selection autogenerated6.88low · n=2$0.023179no
Relevance, Classification & Matchingx_post_selection autogenerated7.79medium · n=100$0.00322880%no
Social & Promotional Contentauto_reddit_post_generation autogenerated7.49high · n=100$0.061540-23%no
Social & Promotional ContentSocial Post Promo (pooled)7.39ranked · n=94$0.003608-2553%no
Social & Promotional Contentx_com_messages_for_promotion autogenerated0.00low · n=0$0.032237no
Content Summarization & Synthesiscontent-summarization autogenerated6.47ranked · n=100$0.008622-320%no
Content Summarization & Synthesisexecutive_summary_generation autogenerated0.00low · n=0$0.035708no
Content Summarization & Synthesissynthesis_of_titles_for_publication autogenerated8.01high · n=66$0.00600940%no
Topic Organization & Clusteringps_section_reassignment autogenerated0.00low · n=0$0.008722no
Topic Organization & ClusteringTopic Discovery Clustering (pooled)7.13high · n=100$0.06103480%no
Topic Organization & Clusteringtopic_cluster_naming autogenerated8.22high · n=72$0.00679281%
Topic Organization & Clusteringtopic_sequence_determination autogenerated0.00low · n=0$0.006870no