TrueForge claims 75% cheaper agent runs

TrueFoundry open-sourced the MIT-licensed TrueForge agent harness after reporting that it solved roughly 11 of 14 Enterprise-Bench tasks for $2.90 per run with GLM-5.2, versus about $11.80 for Claude Managed Agents running Opus 4.8. The company’s reproducible benchmark held the 14 tasks, MCP servers, and blind judge constant; with Opus fixed across harnesses, TrueForge reported $8.50 per run versus $11.80 for Claude Managed Agents. The result is a vendor claim, not an independent replication, but teams studying harness-level coding-agent cost have another reminder that model pricing is only one line on the agent bill.