GILD verdict
Choose Codex if OpenAI's models cover the work and you want the agent inside ChatGPT Business at US$20 per user per month billed annually.
Choose Factory if you want several model vendors behind one agent, GitLab review, or a deployment inside your own network. Pilot both on the same repositories and compare merged changes, reviewer time, and spend at the plan you would buy.
Buy on model breadth and procurement, because on Terminal-Bench 2.0 the two agents score within 2 points of each other on the same model. Codex comes inside ChatGPT plans on OpenAI's models, and Factory sells its own plans across several model vendors. Both give an engineer a terminal agent, IDE support, headless runs in CI, cloud work, and AGENTS.md instructions.
Choosing between Factory and Codex
Choose Factory if
- You want Claude, Gemini, Grok, and open-weight models alongside GPT under one plan, or your own keys in the CLI.
- You need agent sandboxes in your own VPC, or a Custom Deployment without a Factory-hosted control plane.
- You want automated reviews on GitLab merge requests as well as GitHub pull requests.
Choose Codex if
- Codex comes inside the ChatGPT Business seat at US$20 per user per month billed annually, while Factory Teams costs US$60 a month plus US$40 per seat.
- OpenAI's models cover your team's work, and you want one vendor for chat and coding.
- You want cloud tasks in OpenAI-managed environments started from ChatGPT on the web or mobile.
Side by side
| Dimension | Factory | Codex |
|---|---|---|
| Where it runs | Droid CLI, desktop app for macOS and Windows, web and mobile, IDEs, droid exec in CI, and Droid Computers | ChatGPT desktop app, Codex CLI, IDE extension, and Codex Cloud |
| Individual priceList prices in US dollars. Included usage differs by plan and is not directly comparable. | Pro: US$20 per month | Included in ChatGPT Plus: US$20 per month |
| Team price | Teams: US$60 per month plus US$40 per seat, up to 10 seats | Included in ChatGPT Business: US$25 per user per month billed monthly, or US$20 billed annually (2+ users) |
| Beyond the plan | Free Droid Core fallback after a rate limit, or prepaid Extra Usage from US$10 | ChatGPT credits, or an API key billed at API rates |
| Models | Anthropic, OpenAI, Google, xAI, and Droid Core open-weight models; GPT-6.1 Sol runs at 0.8× | OpenAI models, including GPT-6.1 Sol and GPT-6 Luna |
| Headless and CI | droid exec, read-only by default until --auto sets an autonomy level; Droid Action for GitHub Actions and GitLab CI | codex exec and the Codex GitHub Action |
| Enterprise deployment | Factory Managed, Factory Private with sandboxes in your VPC through OpenCompute, the EU deployment, or a Custom Deployment when policy rules out a Factory-hosted control plane | Codex Cloud in OpenAI-managed environments, with per-environment internet access settings |
| Terminal-Bench 2.0 on GPT-5.2Stanford University and Laude Institute run Terminal-Bench, and the two results sit inside each other's margin of error. Each agent's best entry: Codex CLI 82.2% with GPT-5.5 (April 23, 2026), Droid 77.3% with GPT-5.3-Codex (February 24, 2026). | 64.9% ±2.8, Droid, entered December 24, 202519 | 62.9% ±3.0, Codex CLI, entered December 18, 202519 |
| Pull request review | Automated review on GitHub pull requests and GitLab merge requests through Droid Action | @codex review on connected GitHub repositories |
What practitioners report, side by side
Each product's figures come from its own practitioners, for the group you pick.
Overall signal
All practitioners rate Codex +44, 22 points above Factory at +22. Most of the gap is in complaints, with 14 of Factory's 54 sources negative against 12 of Codex's 106.
The score counts the authors from all practitioners that take a side, one vote each with their latest view, and is how far positive votes lead negative ones, per 100 votes, so Factory's +22 rests on votes split 24 positive, 13 mixed and 13 negative and Codex's +44 on 55, 30 and 12.
A lead of a fifth of the votes reads positive and a lead of half strong positive, with the same lines below zero. The dots show each source, positive, mixed, negative. Use a task pilot to measure accepted work and review time.
Footprint on the Agentic Software Development stack
They differ on two of 12 layers. On Codebase context, Factory supports it and Codex is core. On Application security, Factory supports it and Codex touches it. Coverage is each product's placement on this stack, as its product page shows it. Core is a job the product owns; Supports and Touches are jobs it helps with. Coverage says nothing about how well a product does them.
| Layer | Factory | Codex |
|---|---|---|
| Build | ||
| Codebase context | Supports | Core |
| Validate | ||
| Application security | Supports | Touches |
The 10 layers where they match
| Layer | Factory | Codex |
|---|---|---|
| Plan | ||
| Work planning | Supports | Supports |
| Build | ||
| Coding | Core | Core |
| Agent runtime | Core | Core |
| Tool execution | Core | Core |
| Validate | ||
| Debugging | Supports | Supports |
| Testing | Supports | Supports |
| Code review | Supports | Supports |
| Ship | ||
| Source control | Touches | Touches |
| CI/CD | Touches | Touches |
| Operate | ||
| Observability | Touches | Touches |
What practitioners praise
Among all practitioners, Codex alone has seven praise themes that reach 5 sources from 3 authors. Code Quality and Pull Request Review lead, at 27 of 33 and 12 of 12. Both carry Developer experience. It is praised in 9 of the 21 Factory sources that discuss it, against 20 of 36 for Codex. Factory alone has three praise themes that reach 5 sources from 3 authors. Model choice and End-to-end execution lead, at 15 of 22 and 9 of 9.
| Praised for | Factory | Codex |
|---|---|---|
| Developer experience | 9 of 21 | 20 of 36 |
| Model choice | 15 of 22 | Not a theme in its sources |
| End-to-end execution | 9 of 9 | Not a theme in its sources |
| Background & delegated work | 7 of 20 | Not a theme in its sources |
| Code Quality | Not a theme in its sources | 27 of 33 |
| Pull Request Review | Not a theme in its sources | 12 of 12 |
| Long-Running Autonomy | Not a theme in its sources | 12 of 17 |
| Value for Money | Not a theme in its sources | 11 of 15 |
| Codebase Understanding | Not a theme in its sources | 10 of 11 |
| Fast Runs | Not a theme in its sources | 6 of 12 |
| Stability | Not a theme in its sources | 1 of 6 |
Each count is the sources that praise a theme, or raise it as a concern, out of those that discuss it. A row pairs the two products when both carry a theme under the same or an equivalent name.
A count ending in "so far" has under 5 sources from 3 authors among all practitioners. "Not a theme in its sources" means GILD found no theme of that name in that product's sources.
What practitioners raise as concerns
Among all practitioners, Codex alone has five concern themes that reach 5 sources from 3 authors. Inconsistent Results and Speed lead, at 15 of 16 and 13 of 22. Both carry Usage and cost. It is raised in 8 of the 15 Factory sources that discuss it, against 22 of 39 for Codex. Factory alone has three concern themes that reach 5 sources from 3 authors. Reliability & bugs and Permissions & safety controls lead, at 28 of 28 and 11 of 15.
| Raised as a concern | Factory | Codex |
|---|---|---|
| Usage and cost | 8 of 15 | 22 of 39 |
| Reliability & bugs | 28 of 28 | Not a theme in its sources |
| Permissions & safety controls | 11 of 15 | Not a theme in its sources |
| Context & compaction | 7 of 7 | Not a theme in its sources |
| Inconsistent Results | Not a theme in its sources | 15 of 16 |
| Speed | Not a theme in its sources | 13 of 22 |
| Requires Strong Review | Not a theme in its sources | 9 of 23 |
| Communication & Steering | Not a theme in its sources | 7 of 12 |
| Needs Close Steering | Not a theme in its sources | 6 of 9 |
19 sources
- Factory docs: Individual Plans
- Factory docs: Organization Plans
- Factory docs: Available Models
- Factory docs: Droid CLI
- Factory docs: IDE Integrations
- Factory docs: Droid Exec (Headless)
- Factory docs: Automated Code Review
- Factory docs: Deployment Patterns
- Factory docs: AGENTS.md
- Codex docs: Pricing
- Codex docs: Models
- Codex docs: IDE extension
- Codex docs: Cloud
- Codex docs: Cloud environments
- Codex docs: AGENTS.md
- Codex docs: GitHub code review
- Codex docs: Non-interactive mode
- Codex docs: Codex GitHub Action
- Snorkel AI: Terminal-Bench 2.0 leaderboard, Stanford University and Laude Institute (as of Oct 25, 2026)