ai economics
Meta Agent Costs Up to $50K per Million Messages
Meta now bills WhatsApp Business Agent by tokens, turning one million typical delivered messages into a $40,000–$50,000 decision.
The Weighted Average
Cost per 1M typical delivered WhatsApp agent messages $40K–$50K — Meta Agent Costs Up to $50K per Million Messages Monthly stream traffic removed in 100-session scenario 28.5TB — Agent Browser Fixes the Queue Before the Model Time removed from 512-token prefill at 30K context 4.92s — llama.cpp Cuts DeepSeek V4 Prefill by 4.92 Seconds Target active user-days across the 25-seat trial 300 — Copilot in 30 Needs 300 Active User-Days Extra relevant-result equivalents per query versus Google 0.87 — Onton's Search Benchmark Earns a Pilot, Not a Switch
The Wire
agentic engineering
Octobench Finds a 5-Task Harness Swing
The same GLM model solved 24 coding tasks in one harness and 19 in another. Builders should benchmark the whole system.
agentic engineering
Supabase Evals Turns Agent Support Into Tests
Supabase's live eval gives Codex 100%, but its real lesson is a containerized regression loop for agents, skills, and docs.
Quoted
“The invoice counts delivered messages, but the business case lives or dies on resolved customers.”
About the Paper
“I’m Stephen, a software engineer with over two decades of industry experience. The Weighted Average is a daily reading of the AI economy — the models, the money, and what they do to us.”