Codex cloud tasks vs Factory AI
Read from each vendor’s own pages last night, so it does not go stale the way a written comparison does.
Cloud agent
Codex cloud tasks
Codex's hosted task runner, distributed through ChatGPT rather than sold separately.
Not read
Cloud agent
Factory AI
Enterprise "Droids" for delegated engineering work. No public revenue figures.
$20 / mo
Measured — counted from real repositories, nightlyAsserted — read from the vendor docs, dated
| Compared on | Codex cloud tasks | Factory AI |
|---|---|---|
| Asserted — what each vendor publishes about itself | ||
| Category | Cloud agent | Cloud agent |
| Vendor | OpenAI | Factory |
| Status | Active | Active |
| How it charges | Subscription | Subscription |
| Entry price | Not read | $20 / mo |
| Plans | No named tier on OpenAI's pricing page could be read with a price beside it. |
|
| Default model | openai/gpt-5.5 | None — you choose |
| Your own key | No — bundled only | No — bundled only |
| Source | Proprietary | Proprietary |
| Measured — counted on the last nightly run | ||
| Stars | not measured | not measured |
| Maintenance | not measured | not measured |
| Days since last push | not measured | not measured |
| Top model in it | OpenAI: GPT-5.5 · $11.25 | not measured |
| Its volume | not measured | not measured |
What has changed on either of them
1 entry| What changed | Kind | Was, and is now | Confidence | Occurred |
|---|---|---|---|---|
| Factory AI changed the quota wording on its pricing pageFactory AI | Quota change | limits 150 seatslimits 10x the usage, 150 seats, 5x the usage, unlimited team members | low |
