---
schema_version: "newruntime-topic-hub-v0.1"
type: "topic_hub"
slug: "agent-economics"
title: "Agent economics - New Runtime"
description: "How AI work is priced, measured, routed, and justified by completed task value rather than raw token volume."
answer: ["Agent economics moves measurement from token use to useful completed work.","Cost only makes sense beside success rate, review time, retry loops, latency, and human attention.","The mature metric is cost per accepted outcome, not cost per generation."]
search_intents: ["agent economics","AI coding cost","cost per completed AI task"]
status: "featured"
last_updated: "2026-07-22"
counts: {"total":8,"signals":6,"patterns":1,"posts":1,"atlas":0,"sources":9}
routes: {"html":"https://newruntime.com/topics/agent-economics/","markdown":"https://newruntime.com/topics/agent-economics.md","json":"https://newruntime.com/topics/agent-economics.json"}
top_sources: ["https://anthropic.com/news/claude-sonnet-5","https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-6-flash-3-5-flash-lite-3-5-flash-cyber/","https://claude.com/blog/claude-model-and-effort-level-in-claude-code","https://code.claude.com/docs/en/model-config","https://cognition.com/blog/making-fable-cheaper-than-opus","https://databricks.com/blog/benchmarking-coding-agents-databricks-multi-million-line-codebase","https://engineering.ramp.com/post/ai-spend-value","https://github.com/steipete/agent-scripts/blob/main/skills/codex-first/SKILL.md","https://platform.claude.com/docs/en/agents-and-tools/tool-use/advisor-tool"]
---

# Agent economics - New Runtime

How AI work is priced, measured, routed, and justified by completed task value rather than raw token volume.

## Short Answer

- Agent economics moves measurement from token use to useful completed work.
- Cost only makes sense beside success rate, review time, retry loops, latency, and human attention.
- The mature metric is cost per accepted outcome, not cost per generation.

## Patterns

- [Agent economics moves to completed work](https://newruntime.com/patterns/agent-economics-moves-to-completed-work/): The economically meaningful unit for agent systems is becoming cost per verified completed task rather than cost per token or model call.

## Field Notes

- [Gemini 3.6 Flash Moves the Agent Race Toward Cost per Task](https://newruntime.com/posts/gemini-3-6-flash-agent-efficiency/): Google's Gemini 3.6 Flash release frames the model race around token efficiency, built-in computer use, and specialized cyber agents rather than raw chat intelligence alone.

## Recent Raw Signals

- 2026-07-15: [A pricier model can be cheaper per completed task](https://newruntime.com/signals/fable-cost-per-completed-task/)
- 2026-07-13: [Claude Code separates model choice from effort](https://newruntime.com/signals/claude-code-model-and-effort-are-separate-controls/)
- 2026-07-12: [Databricks benchmarks coding agents on its own codebase](https://newruntime.com/signals/databricks-benchmarks-real-coding-agent-economics/)
- 2026-07-09: [An advisor model can guide a cheaper executor](https://newruntime.com/signals/advisor-model-guides-cheaper-executor/)
- 2026-07-01: [Sonnet moves agent capability down the price curve](https://newruntime.com/signals/sonnet-moves-agent-capability-downmarket/)
- 2026-06-27: [AI spend should be measured per successful task](https://newruntime.com/signals/ai-spend-measured-per-successful-task/)

## Public Sources

- https://anthropic.com/news/claude-sonnet-5
- https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-6-flash-3-5-flash-lite-3-5-flash-cyber/
- https://claude.com/blog/claude-model-and-effort-level-in-claude-code
- https://code.claude.com/docs/en/model-config
- https://cognition.com/blog/making-fable-cheaper-than-opus
- https://databricks.com/blog/benchmarking-coding-agents-databricks-multi-million-line-codebase
- https://engineering.ramp.com/post/ai-spend-value
- https://github.com/steipete/agent-scripts/blob/main/skills/codex-first/SKILL.md
- https://platform.claude.com/docs/en/agents-and-tools/tool-use/advisor-tool