The Setuproll Leaderboard

Benchmarks

74Builds Ranked
96.0Top Score
87.0Avg Score
$0.06Cheapest

Cost vs Score

top-left wins
707580859095$0.00$0.27$0.54Cost per taskScore

Each dot is a build. Up means higher score, left means cheaper. Hover for details, click to open.

Full Leaderboard

click a column to sort
#BuildModel
1SAlex Tern's Full-Stack Claude RigClaude CodeClaude Code logoClaude Opus 4.8 (plan + review)96.091%2.8s$0.522.5k
2SClaude Code Full-Stack PrimeClaude CodeClaude Code logoClaude Opus 4.894.089%2.1s$0.461.3k
3SClaude Telegram Bot With Long-Term MemoryClaude CodeClaude Code logoClaude Sonnet 4.693.091%3.2s$0.512.8k
4SCut a podcast into viral shorts and auto-postClaude CodeClaude Code logoClaude Opus 4.893.092%2.7s$0.542.8k
5SBest Setup to Build a REST API with Postgres and TestsClaude CodeClaude Code logoClaude Opus 4.893.089%2.6s$0.512.7k
6SServe a Private OpenAI-Compatible API with vLLMAiderAider logoLlama 3.3 70B Instruct93.090%2.9s$0.512.5k
7SAuto-Triage Your Inbox and Draft Replies (n8n + Claude)Claude CodeClaude Code logoClaude Sonnet 4.6 (classify)92.090%2.5s$0.502.8k
8SIdea to Deployed SaaS in a Weekend (Lovable + Claude Code)Claude CodeClaude Code logoClaude Opus 4.8 (logic)92.086%2.5s$0.532.8k
9SDiscord community bot with Claude and long-term memoryClaude CodeClaude Code logoClaude Sonnet 4.692.092%2.1s$0.442.1k
10SAI Data Analyst: Jupyter + Postgres MCP Notebook LabClaude CodeClaude Code logoClaude Opus 4.892.087%2.3s$0.461.7k
11SClaude Code Backend API ForgeClaude CodeClaude Code logoClaude Opus 4.892.087%2.7s$0.541.0k
12SClaude Code + Playwright QA RigClaude CodeClaude Code logoClaude Sonnet 4.692.087%3.1s$0.50877
13SRun Flux locally on a Mac with ComfyUI for unlimited free rendersClaude CodeClaude Code logoFlux.1 dev (fp8)91.091%2.6s$0.523.1k
14SBest Local LLM on a Mac (Apple Silicon)Claude CodeClaude Code logoQwen3 32B (Q4_K_M)91.091%1.8s$0.443.1k
15SFaceless Channel: Script to Upload in One SittingClaude CodeClaude Code logoClaude Opus 4.8 (outline + script)91.085%2.7s$0.522.8k
16STelegram RAG Bot Over Your Own DocsClaude CodeClaude Code logoClaude Opus 4.891.088%2.7s$0.511.5k
17ACursor Max-Quality OpusCursorCursor logoClaude Opus 4.891.085%2.8s$0.481.4k
18SSlack RAG assistant that searches your company wikiClaude CodeClaude Code logoClaude Opus 4.890.084%2.8s$0.522.5k
19SFastest Frontend From a Figma Design (Cursor)CursorCursor logoClaude Sonnet 4.690.086%2.7s$0.432.2k
20SCursor Tab VelocityCursorCursor logoClaude Sonnet 4.690.086%2.2s$0.452.1k
21SLLM Eval Harness: Score Prompts Before You ShipClaude CodeClaude Code logoClaude Sonnet 4.6 (under test)90.084%2.4s$0.441.3k
22SFully automated clip-and-post pipeline with n8nClaude CodeClaude Code logoClaude Opus 4.890.084%2.6s$0.491.2k
23SClaude Code Legacy Refactor EngineClaude CodeClaude Code logoClaude Opus 4.890.086%2.0s$0.51689
24AClaude Code Monorepo OrchestraClaude CodeClaude Code logoClaude Opus 4.890.085%1.9s$0.48506
25SCut a Podcast Into 10 Viral Shorts and Auto-PostClaude CodeClaude Code logoGemini 2.5 Pro (transcript scan + moment ranking)89.087%2.9s$0.453.4k
26SLock a consistent brand style in Midjourney with style referencesCursorCursor logoMidjourney v789.086%2.7s$0.532.7k
27ACursor Frontend Figma PipelineCursorCursor logoClaude Sonnet 4.689.087%2.3s$0.411.2k
28AClaude Code Security Audit CellClaude CodeClaude Code logoClaude Opus 4.889.083%2.7s$0.38558
29ANo-Code GPT-5 Telegram Bot In n8nCursorCursor logoGPT-588.088%1.9s$0.433.1k
30ATrain a Flux LoRA of one character and reuse it across every sceneClaude CodeClaude Code logoFlux.1 dev base88.084%2.2s$0.441.9k
31AAuto-Enrich and Score New Leads into Your CRM (Make)CursorCursor logoGPT-5 (scoring + pitch angle)88.082%2.5s$0.391.7k
32AWhatsApp support bot with n8n and a knowledge baseCursorCursor logoGPT-588.084%2.8s$0.431.7k
33ARepurpose a webinar into talking-head clips with DescriptCursorCursor logoClaude Sonnet 4.688.087%1.9s$0.361.6k
34ASetup for a Type-Safe GraphQL API End to EndCursorCursor logoClaude Sonnet 4.688.085%2.3s$0.451.1k
35ACursor PR Review CompanionCursorCursor logoClaude Opus 4.888.081%2.7s$0.40821
36AClaude Code DevOps IaC PilotClaude CodeClaude Code logoClaude Opus 4.888.083%2.6s$0.41734
37AClaude Code Data/ML Notebook LabClaude CodeClaude Code logoClaude Opus 4.888.087%2.0s$0.39612
38ASQL Analytics Copilot for Warehouse QuestionsCursorCursor logoClaude Sonnet 4.687.085%2.3s$0.391.5k
39ASetup to Containerize and Deploy a Backend ServiceClaude CodeClaude Code logoClaude Opus 4.887.084%2.1s$0.451.3k
40ADaily Competitor and Topic Research Agent (CrewAI + n8n)Claude CodeClaude Code logoClaude Sonnet 4.6 (searchers)87.081%2.1s$0.461.3k
41AClaude Code Budget Sonnet 1MClaude CodeClaude Code logoClaude Sonnet 4.687.085%2.4s$0.39942
42APrivate Local RAG Over Your Own DocumentsContinueContinue logoQwen3 32B (generation)86.083%1.5s$0.412.0k
43ACheapest Reliable Setup for Everyday CRUD APIsClaude CodeClaude Code logoClaude Sonnet 4.686.080%1.8s$0.422.0k
44AFully Local Private Telegram Bot With OllamaClaude CodeClaude Code logoLlama 3.x 70B86.081%2.4s$0.452.0k
45AThumbnail and Title A/B Lab That Beats Your CTRCursorCursor logoGPT-5 (concept + title brainstorming)86.086%2.3s$0.381.9k
46ALocal Coding Copilot in VS Code (Zero Cloud)ContinueContinue logoQwen3 Coder 30B (chat)86.083%1.9s$0.471.8k
47AGenerate thousands of product images via the Flux API on ReplicateCursorCursor logoFlux.1 pro (Replicate)86.082%2.2s$0.431.5k
48ABest Setup for a Marketing Landing Page (v0)CursorCursor logov0 (GPT-5 backed)86.084%1.5s$0.421.5k
49AClaude Code Prototype SprintClaude CodeClaude Code logoClaude Sonnet 4.686.083%1.7s$0.471.4k
50APremium shorts with AI B-roll for brand accountsClaude CodeClaude Code logoClaude Opus 4.886.086%2.6s$0.47740
51ACopilot Enterprise GuardrailsGitHub CopilotGitHub Copilot logoGPT-5.3 Codex85.083%1.6s$0.451.8k
52AAI Support Agent That Answers from Your Docs (Zapier + RAG)CursorCursor logoGPT-5 (answer)85.082%1.8s$0.381.5k
53AFaceless TikTok factory from one long videoCursorCursor logoGemini 2.5 Pro85.078%1.5s$0.401.3k
54ADiscord image-generation bot powered by ComfyUICursorCursor logoFlux + ComfyUI workflow85.082%1.4s$0.441.1k
55ACheapest Way to Validate an App Idea in an Hour (Bolt)CursorCursor logoClaude Sonnet 4.684.078%1.8s$0.421.7k
56ALocal Reasoning Model on a WorkstationClaude CodeClaude Code logoDeepSeek-R1 Distill Qwen 32B84.081%1.5s$0.461.6k
57AMultimodal Telegram Bot: Voice, Photos, CommandsCursorCursor logoGemini 2.5 Pro84.081%1.3s$0.461.2k
58ADeep Research to Storyboard for Documentary VideosClaude CodeClaude Code logoClaude Opus 4.8 (research + narrative)84.077%1.8s$0.371.1k
59AFine-Tune an Open Model on One GPU (QLoRA)AiderAider logoLlama 3.1 8B84.082%1.9s$0.40980
60BSetup to Add Stripe Billing and Webhooks SafelyCursorCursor logoClaude Opus 4.884.076%1.8s$0.19960
61BClaude Code Mobile RN StudioClaude CodeClaude Code logoClaude Sonnet 4.684.081%1.5s$0.20521
62BAuto-generate blog and social thumbnails with n8n and IdeogramClaude CodeClaude Code logoGPT-5 (prompt writer)83.077%1.4s$0.261.2k
63BCline VS Code AutonomyClineCline logoClaude Sonnet 4.683.079%1.7s$0.191.1k
64BClaude Code Docs & AEO WriterClaude CodeClaude Code logoClaude Sonnet 4.683.075%1.9s$0.18394
65BWindsurf Data Pipeline CascadeWindsurfWindsurf logoClaude Sonnet 4.682.080%1.7s$0.17437
66BFully Local No-Cloud Automation Agent (n8n + Ollama)Claude CodeClaude Code logoQwen3 32B (Q4_K_M)81.078%2.2s$0.17960
67BWindsurf Cascade OnboardWindsurfWindsurf logoClaude Sonnet 4.681.081%1.8s$0.22763
68BPrivate Local Script Writer on Your Own MachineContinueContinue logoLlama 3.x 70B81.081%1.5s$0.24760
69BAider BYOK TerminalAiderAider logoGemini 2.5 Pro80.073%1.1s$0.191.5k
70BBuild a Web App With No Vendor Lock-In (Cline + Local Model)ClineCline logoClaude Sonnet 4.6 (BYOK)80.076%2.1s$0.231.1k
71BCopilot Solo Cheap TierGitHub CopilotGitHub Copilot logoGPT-5.3 Codex79.075%1.8s$0.191.7k
72BPrivate Viber bot on a self-hosted local LLMAiderAider logoLlama 3.3 70B (Ollama)79.074%1.2s$0.19640
73BRun an LLM on a Cheap CPU-Only BoxAiderAider logoGemma 3 12B78.073%1.7s$0.271.3k
74CAider Cheapest Gemini FlashAiderAider logoGemini 2.5 Flash72.066%1.2s$0.06884

Tiers: S A B C. Speed and cost are mean per task; lower is better.