Microsoft Research plugin TokenEconomizer cuts AI coding agent costs
Start here
Microsoft Research, with UC Berkeley and Microsoft Security, released TokenEconomizer, an open-source plugin that reduced costs by 18.4–30.9% across five AI coding agent and model combinations on 277 SWE-Bench tasks from six repositories, while task success rates stayed broadly stable. The plugin learns from an agent's past sessions in a repository through three mechanisms: micro-skills that add one-line repository notes, pruning of unused tools, and failure-loop intervention that suggests a different approach when a judge model finds the agent stuck. Savings also reached 16.0–21.2% on Web-Bench and 6.4–19.6% on security auditing, and cumulative costs fell below the default in every repository by the sixth task.
Generated from reports · updated 22 min ago
Report timeline
Follow the reports to see the event from different sides.
- Microsoft ResearchPicksTokenEconomizer: Microsoft Research plugin cuts AI coding agent costs 18.4–30.9% on SWE-Bench
TokenEconomizer, an open-source plugin from UC Berkeley, Microsoft Research and Microsoft Security, reduced costs by 18.4–30.9% across five AI coding agent and model combinations on 277 SWE-Bench tasks from six repositories, with task success rates broadly stable. It learns from an agent's past sessions in a repository through three mechanisms: micro-skills that add one-line repository notes, pruning of unused tools, and failure-loop intervention that suggests a different approach when a judge model finds the agent stuck. Savings also reached 16.0–21.2% on Web-Bench and 6.4–19.6% on security auditing, and cumulative costs fell below the default in every repository by the sixth task.
Heat for this event
Not enough continuous observations yet to draw a trend.