Reported News
Factory releases Droid 1.0, an agent harness that runs hundreds of tool calls per model request
Factory says Droid 1.0, released October 29, 2026, is "more accurate, efficient, and faster than ever before". In Factory's own Terminal-Bench 4 runs with GPT-6.1 Sol, Droid passed 57.6% of tasks at $0.79 per passed task, against 54.0% for Codex.
- Changes reported
- 6
- listed below, each from a source we read
- We tested
- 0
- we did not run it ourselves
- Gaps we flag
- 3
- things we could not verify
- Sources
- 2
- 2 primary, linked below
What changed
| # | What the source says | Status |
|---|---|---|
| 01 | Droid 1.0 adds Script mode: the model writes JavaScript that runs in a sandbox and can make hundreds of tool calls in one model request, instead of one model call per tool call. | Reported |
| 02 | Factory says Script mode is on by default for GPT-6.1 Sol, GPT-6.1 Terra, GPT-6 Astra, Claude Fable 5.1, Claude Opus 5.5, Kimi K3 and several older models. | Reported |
| 03 | Factory's own Terminal-Bench 4 numbers: with Claude Opus 5.5, Droid passed 65.2% at $4.73 per passed task versus 57.6% for Claude Code; with GPT-6.1 Sol, 57.6% versus 54.0% for Codex. | Reported |
| 04 | Factory says the app opens faster (first paint 1,839 to 488 ms) and quits faster (median 648 to 30 ms). | Reported |
| 05 | Droid 1.0 ships as a desktop app for macOS and Windows and a terminal UI, plus the Droid SDK (@factory/droid-sdk), which runs the same Droid as the CLI from Node. | Reported |
| 06 | Factory's changelog lists CLI v1.0.0 and Desktop v1.0.0 on October 29, 2026, followed by v1.1.0 on October 30. | Reported |
reported by the sourcemeasured or run by usnot verified
Primary source: Factory: Droid 1.0, building a frontier harness. We are the fast, independent index; the vendor’s own page is the source of record for this release.
In the source’s words
“Droid is our agent harness.”
“A Script can make hundreds of tool calls in one model request”
Early verdict
Worth a look if you run long agent sessions and want one harness across Claude and GPT models, but treat the head-to-head numbers as vendor claims until independent runs appear.
Basis: Based only on Factory's announcement and changelog; we have not run Droid 1.0.
What we have not verified
- All benchmark, cost and speed figures are Factory's own; we did not run Terminal-Bench 4 or compare harnesses ourselves.
- Factory reran some Claude Opus 5.5 trials on Claude Opus 5 after content-moderation stops, per its methodology; we did not check how that affects the comparison.
- We did not install Droid 1.0 or the Droid SDK.
This brief is Reported: written from the sources below, not from running the release ourselves.
Sources
- Factory: Droid 1.0, building a frontier harnessprimary source · fetched 2026-10-31
- Factory changelogprimary source · fetched 2026-10-31
Related
- GPT-6.1 Terra: specs, price and alternatives
- GPT-6.1 Sol: specs, price and alternatives
- Claude Opus 5.5: specs, price and alternatives
- Claude Fable 5.1: specs, price and alternatives
- GPT-6 Astra: specs, price and alternatives
- Kimi K3: specs, price and alternatives
- Claude Opus 5.5 vs GPT-6.1 Sol
- Gemini 4 Argon vs GPT-6.1 Sol
- Claude Sonnet 5.5 vs GPT-6.1 Sol
- Claude Code: verdict and setup guide
- Codex CLI: verdict and setup guide
- opencode: verdict and setup guide
- Cursor: verdict and setup guide
- Devin: verdict and setup guide
- All news briefs
- This week's releases
- New AI models this week
Spotted an error? See our corrections policy or email the address on the contact page. Markdown version for agents: /news/factory-droid-1-0.md
Want these briefs in your Google Top Stories? Add Agentic Weekly as a preferred source on Google.
The Weekly Verdict · free
5 new-model use cases. Every Monday.
A free email every Monday: the week's few big AI releases, then 5 short use cases for the new models, each with a prompt you can paste.
Free. Double opt-in. One click to unsubscribe.