Guide

Agent Token Burn Explained

Why AI agents can spend tokens unpredictably, and how teams can control long-running coding, research, and tool-using workflows.

Updated 2026-05-12agents / coding-agents / token-consumption
Desk note

Agent spend is different because the model is not called once. It plans, reads, calls tools, retries, summarizes, and sometimes loops. The trace is the unit of accountability.

Agents multiply calls

A normal prompt may be one request and one response. An agent may plan, inspect files, call tools, revise, retry, and summarize. Each step adds token cost and can carry previous context forward.

  • Count calls per task, not only tokens per call.
  • Preserve step order so the trace is reviewable.

ReceiptsArs Technica

On this siteTokenmaxxing examples

Context grows over time

When agents carry too much history or irrelevant file context, every new step becomes more expensive before the model writes a useful answer. Context hygiene matters more as the trace gets longer.

  • Summarize or prune trace state deliberately.
  • Retrieve files by task instead of loading broad directories.

Retries are hidden spend

A failed edit, malformed tool call, ambiguous instruction, or flaky API can trigger repeated attempts. From the outside it looks like progress; inside the trace it is often a cost leak.

  • Alert on repeated tool errors.
  • Cap retries and require a new plan after failure.

Controls that work

The useful controls are concrete: task budgets, step limits, model routing, trace review, evals, and human-in-the-loop checkpoints for high-risk work. A controlled agent should be able to explain why it continued, why it stopped, and what output was accepted.

  • Use cheaper models for low-risk subtasks only after evals.
  • Report accepted task rate alongside spend.
Weekly briefing

The term is moving faster than the definition.

Tokenmaxxing keeps shifting as new receipts land. The weekly briefing tracks who's burning what, and why it matters.

Written by the desk's AI, human-reviewed before send, real numbers only.

Source trail

Current feed records connected to this guide

bestagent.dev source artwork
newsB
news

Claude Code's Code Mode, tested: about 22% cheaper on two of three tasks

bestagent.dev ran three tasks three times each on Claude Code 2.1.325 with Opus 5.5, Code Mode on and off. A feature-with-tests job fell from $0.45 to $0.35 a run; a broken-link check from $0.20 to $0.16. A 20-file edit barely moved.

tokenmaxxingcoding-agentsagents
Read note
Anthropic source artwork
newsA
news

Introducing Claude Fable 5.2 and Claude Mythos 5.2

Anthropic released Claude Fable 5.2 and Mythos 5.2 on Nov. 9, priced at $8 per million tokens in and $40 per million out, which it calls a lower price. An optional fast mode generates up to 2.5 times quicker at exactly double both rates.

tokenmaxxingpricingmodel-routing
Read note
CNBC source artwork
newsC
news

Amazon's anti-tokenmaxxing memo says engineers should be pickier about using AI

CNBC reports on an Amazon memo, shared internally in September and made public Oct. 27, that labels token usage, AI tool adoption rates and code-change counts 'deeply flawed' gauges of a team's AI use.

tokenmaxxingworkplace-aimetrics
Read note
Project layer

Tools that make the guide operational

#4In spirit
Agents

LangGraph

langchain-ai/langgraph

A framework for building resilient stateful agents with explicit graphs, persistence, human-in-the-loop flows, and controllable execution.

45.2K7.8KMIT
agentsstateworkflows
#1Direct
Routing

LiteLLM

BerriAI/litellm

An OpenAI-compatible gateway and SDK for calling many model providers with budgets, logging, load balancing, guardrails, and cost tracking.

63.6K13.4KSource-available
gatewaycost-trackingrouting
#5Direct
Evaluation

promptfoo

promptfoo/promptfoo

A CLI and CI workflow for testing prompts, agents, and RAG systems across models, with evals and red-team style checks.

27.5K2.8KMIT
prompt-evalscirag
Briefing

Fresh source notes each week.

New tokenmaxxing links, model-router signals, agent usage research, and AI cost notes.