OpenAI Codex CLI launched in April 2025 as an open-source terminal agent for coding. Since then, it has grown into one of the most popular AI coding tools on the market, reaching 3 million weekly active users and 74,000+ GitHub stars in under a year.
In this article, you'll find the latest numbers, data, and facts about Codex CLI. From GitHub stars and npm downloads to benchmarks and pricing to technical specifications and enterprise adoption.
- 74,468 GitHub stars, 14.0M npm downloads (last 30 days), 3M weekly active users (April 8, 2026)
- Rewritten from TypeScript to Rust in June 2025. 696 total releases, Apache 2.0 license
- GPT-5.3-Codex scores 77.3% on Terminal-Bench and 56.8% on SWE-bench Pro
- OpenAI valued at $852B (March 2026), $24-25B ARR, $122B latest funding round
1. What Is OpenAI Codex CLI?
Codex CLI is OpenAI's open-source AI coding agent for the terminal. You describe what you want in natural language. Codex writes code, creates files, runs commands, and handles version control. It runs directly in your terminal, sandboxed for safety.
The project launched on April 16, 2025, and has evolved rapidly. Here are the key milestones:
Date | Milestone | Details |
|---|---|---|
| Apr. 13, 2025 | Repository created | GitHub repo initialized |
| Apr. 16, 2025 | Open-sourced | Public release under Apache 2.0 |
| May 16, 2025 | Codex Web agent | Cloud-based agent preview (formerly Codex) |
| Jun. 2025 | Rust rewrite | Rewritten from TypeScript to Rust for performance |
| Jun. 2025 | $1M Open Source Fund | OpenAI commits $1M to open-source tooling |
| Jan. 14, 2026 | GPT-5.2-Codex | New specialized coding model |
| Feb. 5, 2026 | GPT-5.3-Codex | 77.3% on Terminal-Bench |
| Feb. 2026 | Desktop app | Desktop application, 1M+ downloads |
| Mar. 5, 2026 | GPT-5.4 | 1M token context, computer use capabilities |
| Apr. 2, 2026 | Token-based pricing | Shift to token-based billing model |
| Apr. 8, 2026 | 3M weekly users | Confirmed by Sam Altman |
From launch to 3 million weekly users in under a year. That is aggressive growth, even by AI standards. The Rust rewrite in June 2025 was a pivotal moment, significantly improving startup time and resource usage.
2. GitHub Stars and npm Downloads
Codex CLI is fully open source under the Apache 2.0 license. Here are the key repository metrics:
- 74,468 GitHub stars (as of April 10, 2026)
- 10,507 forks
- 696 total releases
- ~4.4 million GitHub release downloads
- Language: 95.6% Rust (since June 2025 rewrite)
2.1. npm Download Growth
The npm download numbers tell a clear growth story. From 82,000 downloads in the launch month to 14.53 million in March 2026, a 177x increase:
The jump from February to March 2026 stands out. Downloads nearly tripled in a single month (2.5x). That spike coincides with the GPT-5.4 release and the desktop app launch. Total npm downloads over the past 12 months: 32.8 million.
2.2. GitHub Stars Comparison
Codex CLI has strong numbers, but the competition has moved fast. Here is how the three major open-source AI coding CLIs compare:
gradually.aiClaude Code leads with 112,297 stars. Gemini CLI is close behind at 100,857. Codex CLI sits in third place with 74,468. The gap is notable, but Codex has a different distribution strategy: it is deeply tied into ChatGPT subscriptions, which means many users never interact with the GitHub repo directly.
3. Weekly Active Users
On April 8, 2026, Sam Altman confirmed that Codex CLI had reached 3 million weekly active users. The growth trajectory since the beginning of 2026 has been steep:
gradually.aiFrom 600,000 weekly users at the start of 2026 to 3 million in just over three months. That is a 5x increase. The desktop app release in February 2026 was a major accelerator, making Codex accessible to developers who prefer graphical interfaces over pure terminal workflows.
For context: ChatGPT has 900 million weekly active users overall and 50 million paying subscribers. Codex CLI targets a much narrower audience (developers), but 3 million weekly active users is a significant number for a command-line tool.
4. npm Downloads Comparison
Monthly npm downloads are a useful proxy for developer adoption. Here is how the three CLI tools compare over the last 30 days:
Claude Code dominates with 46.3 million downloads. Codex CLI is in second place at 14.0 million. Gemini CLI trails at 3.2 million. The gap between first and second is roughly 3.3x.
That said, npm downloads don't capture the full picture for Codex. Many users access Codex through the desktop app or through ChatGPT's web interface (the Codex Web agent), neither of which shows up in npm stats.
5. VS Code Extension
Both Codex and Claude Code offer VS Code extensions. The install numbers as of February 2026:
Extension | Installs | Rating |
|---|---|---|
| Claude Code | 5.2M | 4.0/5 |
| Codex | 4.9M | 3.4/5 |
Claude Code leads in both installs and user satisfaction. The rating gap is significant: 4.0 vs. 3.4 out of 5. That matches what the Pragmatic Engineer Survey found: developers generally prefer Claude Code's output quality, even if they use Codex for its integration with OpenAI's ecosystem.
6. Coding Benchmarks
OpenAI has released several specialized models for Codex. The benchmark progression shows steady improvement:
Model | Release | SWE-bench | Terminal-Bench | Context |
|---|---|---|---|---|
| codex-1 | May 2025 | 72.1% (Verified) | N/A | ~256K |
| GPT-5.2-Codex | Jan. 2026 | 56.4% (Pro) | 64.0% | ~256K |
| GPT-5.3-Codex | Feb. 2026 | 56.8% (Pro) | 77.3% | ~256K |
| GPT-5.4 | Mar. 2026 | N/A | 75.1% | 1M |
A few things stand out. GPT-5.3-Codex holds the best Terminal-Bench score at 77.3%. Terminal-Bench tests how well an AI agent handles real-world terminal tasks, end to end. That is a strong result.
GPT-5.4 trades a small amount of Terminal-Bench performance (75.1% vs. 77.3%) for a massive context window increase to 1 million tokens. For large codebase navigation, that trade-off makes sense.
7. Pricing and Costs
Codex CLI is included in all ChatGPT subscription plans. Unlike Claude Code (which requires a separate Claude subscription), Codex comes bundled with ChatGPT:
Plan | Price | Message Limits (per 5 hours) |
|---|---|---|
| ChatGPT Plus | $20/month | 30-150 messages |
| ChatGPT Pro | $200/month | 300-1,500 messages |
| ChatGPT Business | $30/user/month | Team-managed limits |
7.1. API Pricing
For API access, OpenAI offers two options:
Model | Input (per 1M tokens) | Output (per 1M tokens) |
|---|---|---|
| codex-mini-latest | $1.50 | $6.00 |
| GPT-5 | $1.25 | $10.00 |
GPT-5 has a lower input price ($1.25 vs. $1.50) but a higher output price ($10.00 vs. $6.00). For code generation tasks where output tends to be longer than input, codex-mini-latest may be the more economical choice.
7.2. Price Comparison with Competitors
gradually.aiAt the entry level, both Codex (via ChatGPT Plus) and Claude Code (via Claude Pro) cost $20/month. The value proposition differs: ChatGPT Plus includes Codex as one feature among many (DALL-E, browsing, file uploads). Claude Pro is more focused on coding and long-form tasks.
8. Technical Specifications
The technical overview:
Property | Value |
|---|---|
| Context window (CLI) | ~256K tokens |
| Context window (GPT-5.4) | 1M tokens |
| Sandbox (macOS) | Seatbelt |
| Sandbox (Linux) | Landlock + seccomp |
| Platforms | macOS, Linux, Windows (WSL) |
| Installation | npm i -g @openai/codex or brew install --cask codex |
| Language | 95.6% Rust |
| License | Apache 2.0 |
| Supported models | GPT-5.4, GPT-5.3-Codex, codex-mini-latest, o3, o4-mini |
| Total releases | 692 |
The sandbox architecture is worth highlighting. Codex uses Seatbelt on macOS and Landlock + seccomp on Linux to isolate executed code. That means Codex can run commands in your terminal without risking damage to your system. It is a design choice that prioritizes safety over convenience, and it is one of the areas where Codex genuinely differentiates itself.
The Rust rewrite is another notable decision. Moving from TypeScript to Rust in June 2025 significantly reduced startup time and memory usage. At 95.6% Rust, Codex is one of the few AI tools built primarily in a systems language.
9. OpenAI Financials
Codex doesn't exist in isolation. It is part of OpenAI's broader ecosystem. Understanding the parent company's financials provides context for where Codex fits:
gradually.ai$852 billion valuation. $168 billion in total funding. $24-25 billion in annualized revenue. These are staggering numbers, even in the context of the AI industry.
9.1. Latest Funding Round
OpenAI's $122 billion round in March 2026 was the largest venture capital deal in history. The investor breakdown:
Amazon alone invested $50 billion. Nvidia and SoftBank contributed $30 billion each. For comparison, Anthropic's largest round was $30 billion total. OpenAI raised 4x that in a single deal.
9.2. Revenue Comparison
Where does OpenAI stand relative to the competition?
gradually.aiOpenAI leads with roughly $25 billion ARR, but Anthropic at $19 billion is closing the gap fast. The AI coding segment is a key driver for both companies. Codex and Claude Code are competing not just for developer mindshare, but for billions in revenue.
10. Enterprise Adoption
The enterprise numbers are OpenAI's strongest argument for Codex:
- 90% of Fortune 100 companies use OpenAI products (including Codex)
- 6x enterprise growth since January 2026
- 900M weekly active users across ChatGPT (50M paying)
- ~4,000 employees (end of 2025), targeting 8,000 by end of 2026
10.1. Named Enterprise Customers
Customer | Industry | Details |
|---|---|---|
| GitHub | Developer Tools | Codex powering Copilot's agentic features |
| Notion | Productivity | AI assistant integration |
| Cisco | Networking | Enterprise deployment |
| Nvidia | Semiconductors | Development workflow integration |
| Goldman Sachs | Finance | Internal coding assistance |
| Ramp | Fintech | Engineering team adoption |
| DoorDash | Delivery | Development acceleration |
10.2. Codex Security
In March 2026, OpenAI launched Codex Security in preview. The early numbers are significant:
- 1.2 million commits scanned
- 10,561 high-severity findings
That is roughly 1 high-severity security issue per 114 commits. Codex Security automates vulnerability detection in the CI/CD pipeline, a direct competitor to Snyk and SonarQube but powered by LLMs. Enterprise customers care deeply about this. Security tooling is often the deciding factor for large organizations choosing an AI coding platform.
11. Survey Data and Market Position
Two major surveys provide context for Codex's market position:
11.1. Pragmatic Engineer Survey (February 2026)
The most comprehensive AI coding survey to date (~900 developers):
- Claude Code ranked #1 most-used and #1 most-loved AI coding tool
- 95% of developers use AI tools weekly
- Codex ranked behind Claude Code and GitHub Copilot in overall usage
11.2. JetBrains Survey (January 2026)
Metric | Value | Context |
|---|---|---|
| Heard of Codex | 27% | Awareness among surveyed developers |
| Currently using Codex | 3% | Pre-desktop app launch |
| AI coding market size (2026) | $9.5-12.8B | Estimated market value |
| Market CAGR | 22-28% | Annual growth rate |
The JetBrains numbers (3% usage) look low, but they were collected in January 2026, before the desktop app launch and the GPT-5.3-Codex/GPT-5.4 releases. The 3 million weekly active users reported in April 2026 suggest that adoption has accelerated dramatically since then.
For complex coding tasks (multi-file refactoring, architecture decisions, large-scale debugging), Claude Code leads at 44%. Codex / ChatGPT sits at 19%. This is the segment where terminal agents have the most impact, and where Codex has the most room to grow.
12. Competitor Overview
The AI coding CLI market has three major open-source players. Here is a direct comparison:
Feature | Codex CLI | Claude Code | Gemini CLI |
|---|---|---|---|
| GitHub Stars | 74,468 | 112,297 | 100,857 |
| npm Downloads (30d) | 14.0M | 46.3M | 3.2M |
| License | Apache 2.0 | Open Source | Apache 2.0 |
| Language | Rust | TypeScript | TypeScript |
| Context (max) | 1M (GPT-5.4) | 1M (beta) | 1M (standard) |
| Terminal-Bench | 77.3% | 65.4% | N/A |
| SWE-bench Pro | 56.8% | N/A | N/A |
| Sandbox | Seatbelt/Landlock | Seatbelt/Landlock | Docker |
| Included in | ChatGPT plans | Claude plans | Free (API) |
| VS Code Extension | 4.9M installs | 5.2M installs | N/A |
Each tool has clear strengths. Codex leads on Terminal-Bench (77.3%) and benefits from the ChatGPT ecosystem with 900 million weekly users. Claude Code leads in overall developer adoption and satisfaction. Gemini CLI offers the largest free context window at 1 million tokens standard.
For a much deeper comparison, read our article on Claude Code vs. Gemini CLI vs. Codex.
13. Interesting Facts and Records
To close out, here are the most notable facts about Codex CLI:
- Rust rewrite in 2 months: Codex was rewritten from TypeScript to Rust in June 2025, just two months after the initial open-source release. 95.6% of the codebase is now Rust.
- 696 releases in 12 months: That averages out to nearly two releases per day. Few open-source projects maintain this pace.
- 177x download growth: From 82,000 npm downloads in April 2025 to 14.53 million in March 2026.
- 3 million weekly users in under a year: Confirmed by Sam Altman on April 8, 2026.
- 77.3% Terminal-Bench: GPT-5.3-Codex holds the highest Terminal-Bench score among all AI coding models tested.
- $122 billion funding round: OpenAI's March 2026 round was the largest VC deal in history. Amazon invested $50 billion alone.
- Codex Security scanned 1.2 million commits: Found 10,561 high-severity findings in the preview period.
- 90% of Fortune 100: Nearly all Fortune 100 companies use OpenAI products, giving Codex a built-in enterprise distribution channel.
The bottom line: Codex CLI has established itself as a serious contender in the AI coding market. It trails Claude Code in developer adoption and satisfaction, but OpenAI's ecosystem advantages (ChatGPT integration, enterprise relationships, and massive funding) give it a strong foundation for growth.
If you want to compare all three CLI tools in detail, check out our Claude Code vs. Gemini CLI vs. Codex comparison.
