New on The AI Loadout
Recent additions, newest first. Follow along with the RSS feed or the daily newsletter.
- PackOct 15, 2026Solo Founder Pack
Seven agent skills for shipping a SaaS on your own: scope it, build it, launch it, support it.
- PackOct 15, 2026Bot Persona Pack
Six ready made personas you can drop into Claude Code, Cursor, Codex or any chat app.
- PackOct 15, 2026Agent Project Manager
A project manager skill that scores each task and routes it to the cheapest model that can do it, from local Gemma 4 to frontier models.
- PackOct 15, 2026Founder Bundle (both packs)
Both packs for less. Every skill and every persona, with updates.
- PackOct 15, 2026Complete Bundle (all 3 packs)
All three packs for less. Every skill, every persona and the Agent Project Manager, with updates.
- SetupOct 15, 2026Best AI setup for coding
Today's best AI loadout for coding: Claude Code with Claude Opus 5.5, plus ready agent skills and a bot persona. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI assistant to pay for
Which $20 AI plan is worth it? ChatGPT Plus, Claude Pro and Google AI Pro compared by task, with prices. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for writing
Today's best AI loadout for writing: Claude Pro with Claude Sonnet 5.5, plus a voice-keeping editor persona. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for research
Today's best AI loadout for research: Perplexity for cited web answers plus NotebookLM for your own sources. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI model for coding
Today's best AI model for coding: Claude Opus 5.5, with Sonnet 5.5 and GPT-6.1 Sol as value picks. Benchmarks and prices linked. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI coding agent
Today's best AI coding agent: Claude Code with Opus 5.5. Codex, Cursor and Copilot compared with prices and sources. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI for summarizing PDFs and long documents
Today's best AI setup for summarizing PDFs: NotebookLM for grounded answers, Claude for long single files. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI for meeting notes
Today's best AI meeting notes setup: Granola for bot-free notes, with Otter, Fireflies and Fathom compared. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI for Excel and Google Sheets
Today's best AI setup for spreadsheets: Copilot in Excel or Gemini in Sheets, with Claude for complex models. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI for image generation
Today's best AI image generator: ChatGPT Images 2.5, with Nano Banana 2.1 for volume and Midjourney V8.2 for style. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI for building an app without coding
Today's best AI setup for building an app without coding: Lovable, with Bolt and Replit compared. Prices and sources. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for blog posts and SEO content
Today's best AI setup for blog posts and SEO content: Perplexity for sourced research plus Claude for drafting. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for marketing
Today's best AI loadout for marketing: ChatGPT Plus for copy and images, plus a Growth Strategist persona. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for social media posts
Today's best AI setup for social media posts: ChatGPT for posts and images, with a voice-keeping persona. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for academic research and papers
Today's best AI setup for academic research: Elicit to find papers, Consensus and Scite to check evidence, NotebookLM to read. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI setup for data analysis
Today's best AI loadout for data analysis: ChatGPT Plus to run code on your data, with Julius and Claude as alternatives. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI for presentations and slides
Today's best AI setup for presentations: Gamma for fast decks, Claude with the pptx skill for PowerPoint files. Updated Oct 15, 2026.
- SetupOct 15, 2026How to run AI locally
Run AI on your own computer for free with Ollama or LM Studio. Which local model fits 8, 16, 24 or 48 GB, with commands. Updated Oct 15, 2026.
- SetupOct 15, 2026Best AI model API for apps and agents
Which AI API should you build on? Official per-token prices and context windows for every major lab, from $0.10 to $50 per million tokens. Updated Oct 15, 2026.
- ModelOct 15, 2026GPT-6 Astra
OpenAI's most capable model: complex reasoning, coding, computer use, research and document creation.
- ModelOct 15, 2026GPT-6 Luna
Cheap, high-volume work with a clear goal: extraction, classification, structured summaries, subagents.
- ModelOct 15, 2026Claude Fable 5.1
Anthropic's top model for demanding reasoning and long-horizon agentic work, when Opus 5.5 at higher effort still falls short.
- ModelOct 15, 2026Gemini 3.8 Flash
Google's fast, low-cost everyday model, with a free tier in the Gemini API.
- ModelOct 15, 2026Grok 4.7
xAI's flagship for coding, agentic tool use and knowledge work, with built-in web and X search.
- ModelOct 15, 2026Muse Spark 1.3
Meta's hosted flagship, tuned for agentic workflows (multi-step tool, browser and long-horizon tasks), with text, image, video, audio and PDF input.
- ModelOct 15, 2026Mistral Large 4
Mistral's open-weight multimodal flagship for general work, with function calling and agents.
- ModelOct 15, 2026DeepSeek V4 Pro
DeepSeek's strongest API model for reasoning and coding at a low price. No vision.
- ModelOct 15, 2026Qwen3.8 Max
Alibaba's flagship for engineering-scale coding, multi-tool agents and vision.
- ModelOct 15, 2026Kimi K3
Moonshot's open-weight flagship for long-horizon coding and knowledge work.
- ModelOct 15, 2026GLM-5.3
Z.ai's open-weight flagship for agentic coding and repository-scale work. Text only.
- ModelOct 15, 2026Amazon Nova Premier
Amazon's most capable Nova model on Amazon Bedrock, for teams already on AWS that want complex reasoning and agent work billed with their AWS account.
- ModelOct 15, 2026Cohere Command A+
Enterprise agents, RAG, tool use and translation across 48 languages, on your own GPUs or via Cohere.
- Local modelOct 15, 2026Gemma 4 (local)
Best all-round local family: E4B for laptops, 26B A4B for strong reasoning and code on a 24 GB GPU or 32 GB Mac.
- Local modelOct 15, 2026Qwen 3.5 (local)
Wide size range from 0.8B to 122B. The 9B is a strong laptop coder; 27B and 35B for workstation use.
- Local modelOct 15, 2026Qwen 3.8 27B (local)
Newest Qwen open model for coding, research and long agentic tasks on a 24 to 32 GB machine.
- Local modelOct 15, 2026gpt-oss (local)
OpenAI's open-weight reasoning models. 20B runs on a 16 GB GPU or Mac; 120B needs about 80 GB.
- Local modelOct 15, 2026Muse Glimmer 30B (local)
Meta's open model for always-on local agents: tool use, long tasks and failure recovery on a single GPU.
- Local modelOct 15, 2026Llama 3.3 70B (local)
Proven general model if you have 48 GB or more of memory. Older than Muse Glimmer but widely supported.
- Local modelOct 15, 2026Mistral Small 3.2 24B (local)
Solid general assistant with vision and reliable function calling on a 24 GB machine.
- Local modelOct 15, 2026Devstral Small 2 24B (local)
Local coding agent: explores codebases, edits multiple files and drives tools.
- Local modelOct 15, 2026Ministral 3 (local)
Small edge models (3B, 8B, 14B) with vision for low-memory laptops.
- Local modelOct 15, 2026DeepSeek-R1 distills (local)
Step-by-step reasoning and math on modest hardware. The 8B (R1-0528 on Qwen3) fits a 16 GB laptop.
- NewsletterOct 15, 2026AI Loadout Daily: 17 AI setups go live
The AI Loadout launches with 17 task setups. Each one names the best tool, model, agent skills and bot persona for the job, with sources and a dated changelog.
Want tested skills and personas installed in one command? See the AI Loadout packs, from $9.