TWIML podcast: cheaper tokens can still mean a bigger AI bill for agents
samcharrington · x · 2026-11-18
On the latest TWIML AI podcast, host Sam Charrington talks with Sadeepa Wijesekara of HPE about agent-era inference economics.
The core point: falling token prices don't guarantee a smaller bill — a single agent job can require dozens of model calls, so total spend may still grow. The episode breaks down which calls are worth paying a frontier provider for and which can be self-hosted or routed to cheaper models, offering a tiered strategy for optimizing agent inference costs.
More from coding & agent
- Microsoft Ignite LAB663: a hands-on loop for improving agents with Copilot traces and Model Router — lee_stott · 2026-11-18
- Agent tweak is 15% faster at nearly same cost — the model admitted it overweighted raw token counts — dosco · 2026-11-18
- Aware browser bets on feeds, not chat, as the right interface for AI agents — HaijunXia · 2026-11-18
- Why Cheaper Tokens Can Still Mean a Bigger AI Bill: The Agent Call-Stacking Problem — samcharrington · 2026-11-18
- DAWE edits 5,064 of 4B weights to fix harness mismatch, no retraining needed — lateinteraction · 2026-11-18
- Why the 15-minute time limit: bench author says it matches median agent runtime in prod — JoshPurtell · 2026-11-18