AI Pulse · the AI knowledge hub
Know what is happening in AI, and what to do about it.
One place for it all: the daily read on what is actually happening, an always-current index of the 7,051 models and tools worth knowing, and the guides to put them to work in your business.
Free daily email
Get the daily AI Pulse in your inbox
One short read every morning on what is actually happening in AI. Free.
You are in
Your first AI Pulse lands tomorrow morning. Keep an eye on your inbox.
The daily read
Today in AI, Wednesday 30 September
What happened, and the one line on why it matters to a business.
The daily email carries 12 to 18 items. The AI Pulse hub carries the broader set, ranked, with source credibility, counterpoints and receipts on every item. Track a story over time on the timeline, or follow research from paper to product in the research tracker.
Claimed 50-agent ad operation
An operator says they run 50 separate Claude agents, one per media-buying task, replacing a human checklist with granular agents. Single LinkedIn post.
The play
Pilot specialized agents on one repeatable advertising workflow, but require human approval, shared reporting, and limits before expanding beyond experiments.
Claude skill mods any PC game
A new Claude Code skill bundles recon, reverse-engineering, and generative art/3D/audio tools so Claude can mod almost any PC game. It has 635 GitHub.
The play
Watch this as a signal that AI-assisted game content creation is broadening, while reviewing licensing and brand risks before commercial use.
Corral kills runaway agent processes
A tiny open-source tool ensures that no process spawned by a command it runs is left alive when it finishes, even daemons that fork away or processes.
The play
Add process cleanup to agent pilots now, preventing hidden background jobs from creating runaway compute bills or unreliable customer workflows.
Higgsfield crosses $1B ARR in 18 months
Higgsfield's Alex Mashrabov claims 20x year-on-year growth and enterprise expansion behind the milestone, an aggressive AI-native scaling benchmark..
The play
Treat the claimed growth as a market signal, not a benchmark, and study which enterprise use cases support repeatable AI revenue.
Magnitude launches self-tuning agent engine
Inference engine startup shipped a tool that compiles and tunes its own kernels on your device so open models run up to 2x faster than llama.cpp..
The play
Test self-tuning inference on your highest-volume open-model workflow, measuring real latency and device costs before changing production.
Menu-bar app watches your Claude sessions
A lightweight companion app surfaces Claude Code session status from the notch without opening a terminal. Picked up 731 stars in three days, a steady.
The play
Use session monitoring tools for teams running coding agents, so managers can spot stalled work without opening every terminal.
The AI index
The models and tools worth knowing
An always-current, independent index of what AI builders actually use, 7,051 entries and counting. Compare, then decide.
AI Models
1,528Compare models on price, context, and capability.
Explore →MCP Servers
4,234The connectors that plug tools into your agents.
Explore →AI Agents
161Autonomous agents and frameworks that do real work.
Explore →Open Source
797The open tools builders are shipping with.
Explore →AI Apps
300AI-native apps worth knowing about.
Explore →Model leaderboard
The state of AI models: intelligence, price, and speed, side by side.
Compare models →State of AI models
Full leaderboard →| Model | Maker | Context | Price /Mtok | License |
|---|---|---|---|---|
| Claude Opus 4.6 vision | Anthropic | 1M | $5 / $25 | Closed |
| Claude Opus 4 vision | Anthropic | 200K | $15 / $75 | Closed |
| Qwen3 Coder 480B A35B | Qwen | 262K | $0.30 / $1 | Open |
| DeepSeek V3 | DeepSeek | 164K | $0.26 / $1.03 | Open |
| Qwen3 Max | Qwen | 262K | $0.78 / $3.90 | Open |
| Gemini 2.5 Flash vision · audio · video | 1M | $0.30 / $2.50 | Closed | |
| o3 vision | OpenAI | 200K | $2 / $8 | Closed |
| Mistral Small 3.2 24B vision | Mistral | 256K | $0.09 / $0.25 | Open |
| GPT-4o vision | OpenAI | 128K | $2.50 / $10 | Closed |
| DeepSeek V3.2 | DeepSeek | 164K | $0.28 / $0.42 | Open |
| Grok 4.5 vision | xAI | 500K | $2 / $6 | Closed |
| Claude Sonnet 4 vision | Anthropic | 200K | $3 / $15 | Closed |
Live from the Enterprise DNA model index. Prices in USD per million tokens, input / output.
What's making money
How people are actually making money with AI
Real pricing and growth stories from AI-native businesses, the plays worth learning from.
Learn to build it
From reading about AI to running it
Free field guides that take you from the news to a working system in your own business.
Working with Claude
The field guide to getting real work done with Claude.
Get the guide →Working with Claude, for Developers
The developer edition: agents, tooling, and build patterns.
Get the guide →The Second Brain Field Guide
Turn AI into a system that remembers and works for you.
Get the guide →Get the daily read
Subscribe once and the daily AI Pulse lands in your inbox each morning. One short read, free, leave any time.
Want more than a read? The Omni Command Centre is how we put all of this to work inside a business. See what an AI command centre could do for you.
Free daily email
Subscribe to the daily AI Pulse
One short read every morning. Free.
You are in
Your first AI Pulse lands tomorrow morning. Keep an eye on your inbox.