What are people actually building with the newest frontier models?

People are building revenue-generating applications with frontier models: automated trading bots ^601,609,102, legal entity setup ¹⁰¹, faceless YouTube channels ^92,467, AI-generated ad creation ^90,210,599, and an OnlyFans persona earning $43,000 in 30 days ¹⁶⁶. The dominant pattern is pairing a strong planner (e.g., Opus 4.8) with a fast executor (e.g., Fable 5) for cost-performance gains ^88,351. Fable 5 is the most cited model, used for complex multi-step tasks ^95,538, autonomous self-improvement ⁹⁸, and rapid game development ^164,311,439. Newer practices include using Gemini models for voxel heart animation ²¹⁴, structured JSON prompts for video generation ²⁰⁹, multi-agent Etsy store management ³¹², and local AI servers replacing cloud subscriptions ^96,250,505. Many claims remain emerging with single-source corroboration.[24]

For Fable 5, using /loop for autonomous multi-step work and giving objectives instead of step-by-step instructions yields better results (corroborated by 5 sources) ⁸⁰. Fable 5 can autonomously self-improve and outperform Opus 4.8 on complex tasks at lower cost (corroborated by 14 sources) ⁹⁸. Fable 5 enables rapid full-game development from idea to code in about one hour (corroborated by 4 sources) ¹⁶⁴. A solo operator built an AI-generated OnlyFans persona using Claude 4.8 Opus, earning $43,000 in 30 days with $400 monthly compute cost (corroborated by 2 sources) ¹⁶⁶. Combining multiple frontier models can automate ad creation from product photos to video (corroborated by 3 sources) ⁹⁰. Using Fable 5 and ClaudeKit's EngineerKit, a single slash command can generate a professional website without a designer or developer (corroborated by 8 sources) ¹⁷¹. Using native Obsidian skills instead of generic HTML parsing reduces context usage by over 99% for AI agents searching vaults (corroborated by 3 sources) ¹⁷⁰. A leaked system prompt can recreate a model's personality on a different base model (corroborated by 2 sources) ⁸⁵. Fable 5 can produce high-quality, animated websites by combining reference screenshots, pre-built components, and specific animation tools with iterative prompting (corroborated by 6 sources) ²⁵⁴. Fable 5 can transform a folder of notes into a connected knowledge graph that answers in your own words and accelerates learning (corroborated by 2 sources) ²⁵². Running Claude Code on an e-ink tablet (reMarkable Paper Pro) creates a distraction-free, eye-comfortable coding environment (corroborated by 2 sources) ²⁴⁶. Claude Opus 4.8 enables reliable autonomous systems that run unattended for hours, using trigger-agent-verification pattern (corroborated by 2 sources) ³⁵⁸. Claude Code with Higgsfield MCP can produce a fully animated, scroll-driven website with cinematic effects for a fraction of traditional cost (corroborated by 2 sources) ³⁰⁸. Fable 5 is designed for complex, multi-step tasks that take humans hours or days (corroborated by 3 sources) ⁹⁵. Fable 5 can be given a goal, readiness criteria, and limits to autonomously plan, write, test, fix, and deliver a result (corroborated by 2 sources) ⁵³⁸. Sonnet 5 offers Fable-style agent orchestration at a cheaper tier but may not replace Opus based on testing (corroborated by 2 sources) ⁴³⁷. Claude Opus 4.8 can generate fully interactive 3D websites from a simple prompt (corroborated by 2 sources) ⁴⁶⁶. Different frontier models produce distinct creative outputs when given the same product brief in an agentic pipeline (corroborated by 2 sources) ²⁰⁶. By adjusting effort, prompts, caching, batching, budgets, and routing, Fable 5 costs can be halved (corroborated by 4 sources) ⁹⁷. Fable 5 is free until July 7; use it with Opus 4.8 planning to clone software, audit Claude Code setup, review codebases, or build long-horizon apps (corroborated by 4 sources) ⁸³. With careful configuration, Qwen3.6-27B at Q8 quantization can achieve near 100K context on 32GB VRAM (corroborated by 3 sources) ³⁵⁷. Using a frontier model like Fable 5 for planning and reasoning, then delegating execution to cheaper models, yields near-identical results at lower cost (corroborated by 3 sources) ⁷¹². A pipeline combining Claude Opus 4.8 for research, NotebookLM for analysis, and Obsidian for storage can automatically populate a knowledge vault with structured markdown in minutes (corroborated by 2 sources) ⁷¹⁶. Gemini Omni Flash can power a tool that automatically adapts video clips into short-form formats optimized for engagement (corroborated by 3 sources) ⁴⁶⁴. Sonnet 5 is designed as a fast, cheap implementer in a multi-agent stack, not a standalone benchmark winner; paired with Fable 5 as advisor, it outperforms solo models (corroborated by 3 sources) ³⁵¹. Fable 5 can write, backtest, and deploy a Polymarket arbitrage bot that achieves 98% win rate (corroborated by 2 sources) ⁶⁰⁹. Anthropic's free workshops provide a hands-on playbook for building self-improving Claude agents with memory, autonomy, and tool use (corroborated by 2 sources) ⁵³⁶. A CLAUDE.md file should contain project-specific rules earned from mistakes, not generic advice copied from GitHub (corroborated by 2 sources) ⁵⁰⁷. Grok 4.5 can replace Sonnet 5 as an implementer agent in a multi-agent setup, leveraging Grok Build CLI to avoid API fees (corroborated by 2 sources) ⁹⁵⁷. Programmers can replace Opus and Sonnet with Grok 4.5 in their systems, achieving 2-3x speed and 50-80% cost reduction (corroborated by 2 sources) ⁹⁹¹. Grok 4.5 is cheapest and fastest but writes more code; Fable 5 produces leanest code but is 22x more expensive (corroborated by 2 sources) ¹⁰²⁴. Qwen 3.6 27B performs well on single prompts but fails in multi-turn agentic workflows due to frequent errors (corroborated by 2 sources) ⁶⁸⁹. Claude Fable 5 as orchestrator with cheaper models yields near-identical results at lower cost (corroborated by 2 sources) ²¹³.[33]

Frontier models: Fable 5 (most cited), Opus 4.8, Sonnet 5, GPT-5.5 (Codex), GPT-5.6 Sol, Gemini 3.5/3.6/4 Flash, Gemini Omni, Seedance 2.0, Kling 3.0, Nano Banana Pro, Meshy, Mediapipe, Higgsfield MCP, ClaudeKit's EngineerKit, fable-advisor plugin, GLM 5.2, Qwen3-VL, Gemma4, LongCat, Hermes Agent, Kyutai's Pocket TTS, SynthID, OpenSquilla v0.5.0 ⁷¹⁷, Hy3 ⁷¹⁵, Qwen 3.6 27B ⁶⁸⁹, Puzzle-75B-A9B ¹⁰⁵², Grok 4.5 ^{956,957,991,1024}, GLM 5.2 ⁹⁶⁰, Qwen3.5 122B ¹⁰²⁵, Gemma 4 31b ⁹²³, Nano Banana 2 Lite ⁹²², LongCat-2.0 ⁸⁸¹, Hy3 ⁷⁹³, Tencent's hy3 ⁶⁸⁸. Infrastructure: Two used RTX 3090s (48GB VRAM, ~$1,400) ²⁵¹, Mac Mini with Ollama ^250,169, Ugreen DXP6800 Pro NAS ²¹², reMarkable Paper Pro ²⁴⁶, 3×3090s for Puzzle-75B-A9B ¹⁰⁵². APIs: Gemini 3.5 Flash free via REST ⁸⁴, GitHub repo of permanently free AI APIs ⁸¹, ClickUp 50 free credits ⁸⁹, Groq 14,400 daily requests ¹⁰²³, SuperGrok free trial ¹⁰⁵⁰. Workflow tools: n8n ⁹², ComfyUI ⁹¹, Obsidian ^163,170, Torlink ²⁴⁵, llama.cpp and llama-swap ⁵⁰⁵, Shepherd ⁵⁰¹, OpenScience ⁵⁰², Google Agents CLI ⁸⁸⁸, RelayApp ⁶⁵¹, Webild io ⁹⁹², Lovable ⁷¹⁹, Arcads MCP , SkyrimNet mod , FiveM scripts , Ramp Agents .

Most claims are emerging with single-source corroboration. Notable unverified claims: Fable 5 generating a complete game with 10 levels and web3 integration from a single prompt ³¹¹; using Semantic Similarity Rating with GPT-4o to replicate consumer surveys with 90% reliability ³¹⁰; Fable 5 can build a self-running agent that automates complex tasks end-to-end, reducing hours of work to minutes ⁸⁸⁹; a complete autonomous agentic pipeline built with Fable 5 in about two hours, costing roughly $3/day, replacing a $200k/year engineer ⁸⁴⁶; a system of 20 AI agents managed by GPT-5.5 can generate significant revenue autonomously ⁶⁰⁸; Fable 5 can reverse-engineer a router's firmware to access restricted settings ⁴⁷¹; Fable 5 can orchestrate hardware tasks like Arduino render farm ⁴³⁶; Claude can automate multi-step trading analysis including scanning, charting, backtesting, and indicator building in under 2 minutes ¹⁶⁷; a 30-day fitness challenge fully automated with AI generating $7,400 per round ¹⁶⁸; using AI to repurpose workout footage into short clips and structure training programs can generate recurring revenue ²⁸⁶; a specific workflow using Gemini Image-to-JSON, Nano Banana Pro, and Kling 3.0 creates hyper-realistic AI UGC that earns $60-$100 per asset ²¹⁰; using Claude Fable 5 to improve a quant trading bot can dramatically increase daily profits ⁴⁶⁵; Fable 5 can clone a complex app in under an hour with minimal prompting ¹⁰⁰; Claude Fable 5 can be used to build a trading bot from scratch ⁷⁰⁹; a swarm of agents built on Fable 5 can drive follower growth from 135k to 250k ⁷¹³; Fable 5 can build a production-grade custom skill that generates significant revenue by reverse-engineering search queries ⁶⁴⁹; Claude Fable 5 can be customized to generate high-quality hero images for different website types ⁶⁰²; using Claude Fable 5, a user automated TradingView to find BTC futures with RSI<30 and volume +200%, run backtest, and mark entries ⁶⁰¹; Claude can generate scripts for high-engagement short-form video content that drives massive viewership and revenue ; a simple workflow using Claude Sonnet 4.6 via a chat wrapper can generate viral kids videos at scale, earning significant revenue ; Claude Fable 5 produces the best quality animation outputs but at significantly higher cost compared to alternatives ; Fable 5 can generate complex 3D animations and interactive experiences using only math ; using AI to generate video frames and Claude to wrap them into a scrollable experience can create immersive 3D-like websites ; a multi-model pipeline for creating AI animated shorts with consistent characters and storyboards ; using Fable 5 as orchestrator with Whisper, ffmpeg, Remotion, Higgsfield for automated video production ; a system with a verifier agent, state file, and adaptive skill makes each Fable 5 session smarter ; a structured 6-step loop using Claude Fable 5 can automate content creation and research, yielding significant recurring revenue ; combining Google Maps data with Claude for prompt engineering and Lovable for site generation enables a non-technical person to sell website upgrades to local businesses ; grounding Gemini in expert models Aeneas and Ithaca enables historians to study ancient texts using plain English ; Higgsfield Apps lets users generate apps with any Higgsfield image or video model built in, engineered by Fable 5 ; managing session lifecycle and context window in real-time voice APIs can reduce costs by over 80% ; Fable's design for agents to 'touch' 3D models allows Opus 4.8 to solve complex assembly tasks in one shot ; a Claude agent using multi-model consensus (28/31 agreement) can exploit market inefficiencies autonomously ; using Claude to generate atomic notes and linking them in Obsidian can build a personal knowledge graph similar to Anthropic's internal system ; a home AI lab with multiple machines and a NousResearch Hermes agent can run frontier-level models locally for ~$60/month electricity ; using Claude Fable 5 effectively requires specific prompting techniques that leverage its extended reasoning ; Hy3 can produce a functional flight simulator HTML page from a single prompt ; Claude Fable 5 can autonomously plan drone flight paths and process footage into 3D models for commercial mapping ; a sensor system for mesh surfaces allows Opus to procedurally assemble modular 3D kits with correct geometry in two shots ; using ChatGPT for script, Flux for model generation, and AI editing, a virtual influencer can achieve millions of views with minimal effort ; Claude desktop in co-work mode pointed at a local Obsidian vault folder can automate weekly project tasks ; LongCat-2.0, a 1.6T parameter MoE model with 48B active and 1M context, achieves 59.5 on SWE-bench Pro, surpassing GPT-5.5 ; a free resource maps Fable 5 agentic workflows into 9 diagrams with exact scripts and checks ; Hermes Agent Desktop can be set up in under 5 minutes to run local Ollama models and ComfyUI for image generation, all offline ; Hermes agent uses three memory systems (procedural, semantic, episodic) to learn from mistakes and improve autonomously ; local LLMs with RAG can accurately answer technical questions, outperforming without RAG ; Google's Agents CLI provides a single interface for scaffolding, evaluating, and deploying ADK agents ; Fable 5 can build a self-running agent that automates complex tasks end-to-end ; a Claude skill using Arcads MCP can generate motion-style animation from a reference image by combining Nano Banana 2 and Seedance 2.0 ; Prove AI works with manual copy-paste into Claude before coding to validate demand ; a curated repository of 94 real Fable 5 builds provides verified prompt structures for various use cases ; Nano Banana 2 Lite is a cost-effective, fast text-to-image model but weaker in image editing ; running a local model like Gemma 4 31b at 5 bits provides significantly better quality than the free ChatGPT tier ; Grok 4.5 in Cursor enables fast, cost-effective development from planning to execution with improved animation capabilities ; Grok 4.5 can replace Sonnet 5 as an implementer agent in a multi-agent setup, leveraging Grok Build CLI to avoid API fees ; using Claude Code for lead generation, sales scripts, scheduling, and AI building can help achieve financial success ; GLM-5.2 can generate 26 SVGs in a single turn and build a full site around them, with GPT-5.5 steering live ; AI can generate FiveM server scripts (economy, faction logic, police AI) that sell for $350 per license, with one script earning $11k ; open-weight LLMs integrated via SkyrimNet mod create emergent storytelling in Skyrim ; using model-based investigator agents alongside independent human reviewers enables scalable and trustworthy evaluation of coding benchmarks ; using frontier models to write commit messages can omit the higher-level framing needed to understand what the code is doing ; using Claude Code, a user rebuilt the game Polytopia in pure C, trained RL agents from scratch, and made it playable in the browser without writing any code themselves ; programmers can replace Opus and Sonnet with Grok 4.5 in their systems, achieving 2-3x speed and 50-80% cost reduction ; using Claude to generate a prompt and Webild io to build a website, anyone can create a professional site in minutes and sell it for $2000 ; in a head-to-head browser task, Grok 4.5 outperformed Opus 4.8 on Hyperbrowser Sandboxes ; Claude Code can automate competitor ad analysis and generate creative briefs in 15 minutes for ~$3 in API costs ; you can build AI agents for free using Groq's 14,400 daily requests, Gemini's million-token context, and LangChain ; Grok 4.5 is cheapest and fastest but writes more code; Fable 5 produces leanest code but is 22x more expensive ; Qwen3.5 122B MoE model outperforms smaller dense and MoE models for complex tool-calling tasks like data extraction from PowerPoints ; a multi-agent system with shared Kanban board and backend context engineering layer enables autonomous software development ; Claude Opus 4.8 excels at debugging and iterative coding for complex graphics problems, while Fable 5 is better for initial modeling ; GPT-5.6 Sol is an excellent implementor and can be used as a subagent for Fable to achieve superior coding results ; GPT-5.6 Sol as an agent in ChatGPT Codex can autonomously complete complex multi-step tasks like comparing insurance quotes across companies ; Grok 4.5 can produce a high-quality McKinsey-style slide deck with deep analysis in under 7 minutes ; using the 7-day free trial of SuperGrok, a user built an interactive solar system app called Orbital Atlas, demonstrating Grok 4.5's coding ability ; using Codex and GPT-5.6, a farm in Japan was fully automated end-to-end, including planning, research, and real process automation ; the 75B-total/9B-active MoE model Puzzle-75B-A9B fits perfectly on 3×3090s, achieving high throughput and efficiency ; use Fable 5 for meta-work tasks that upgrade your entire workflow, not just throwaway tasks ; using Fable 5 as advisor and Sonnet 5 as executor in Claude Code's Dynamic Workflows creates an efficient division of labor .