Factory releases Droid 1.0 coding-agent harness with script mode
Original title: Meet Factory Droid 1.0: A Model-Independent Coding Agent Harness Built on Script Mode
Programmatic tool calling is now default; Factory reports 65.2% on Terminal-Bench 4 with Claude Opus 5.5.
Factory released Droid 1.0, a model-independent coding agent harness for macOS, Windows, a terminal UI and the Node SDK. The default Script tool lets a model write sandboxed JavaScript that can drive many tool calls while only printed output returns to the model. Factory's own Terminal-Bench 4 runs put Droid ahead of several other harnesses on the same models, with lower reported cost on Opus 5.5. Paid plans start at $20 per month.
Unconfirmed: Terminal-Bench scores and cost figures are Factory's own runs, not independent results.
- Base prompt is 13,055 tokens, down 37% from 20,887.
- Only 2 tool definitions ship up front, down from 11.
- Opus 5.5 score is 65.2% at $4.73 per passed task, self-reported.
- GPT-6.1 Sol score is 57.6%; Codex scored 54.0% in the same comparison.
- Script mode is default for 10 models, including GPT-6.1 Sol and Kimi K3.
- Vendors
- Factory
- Products
- Droid 1.0
- AI models
- Claude Opus 5.5GPT-6.1 SolKimi K3
From the original article
7 min read at marktechpost.com
Factory released Droid 1. 0, a model-independent coding agent harness that makes programmatic tool calling the default. On Factory’s own Terminal-Bench 4 runs, Droid scores 65.
By Asif Razzaq · marktechpost.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.