Codex

Excellent

by OpenAI · founded 2025 · updated Sep 2026

Codex is OpenAI's coding agent: give it a task and it works in a sandboxed environment or your repo, writing code, running tests and opening pull requests. Available in ChatGPT, a CLI and IDE extensions, it represents OpenAI's push from code completion toward delegated, agentic software engineering.

Coding AssistantsSubscriptionEstablishedFrom Included with ChatGPT Plus / Pro
85.3TIP Score
Excellent

The score, taken apart

Fixed weights, sources attached, the formula

Cap 89Val 84Sec 80Int 84Mat 76Mom 92
Capability(30%)89
Value for Money(15%)84
Security & Compliance(15%)80
Integrations & Ecosystem(15%)84
Maturity & Reliability(10%)76
Momentum(15%)92

Best for

  • ◆Delegating well-scoped engineering tasks
  • ◆ChatGPT-subscribed developers
  • ◆Parallel background coding work

Key integrations

GitHubChatGPTCLI & IDE extensionsCloud sandboxes

Strengths

  • +Delegated tasks: writes code, runs tests, opens PRs
  • +Backed by OpenAI's strongest coding models
  • +Parallel task execution in cloud sandboxes

Limitations

  • -Younger than incumbent coding assistants
  • -Best experience requires ChatGPT paid plans
  • -Long autonomous runs still need review

Evidence behind this score

Each score is built from the claims below. This is a snapshot, so check the current terms with the vendor before you commit to anything.

1 of 4 claims below are past their re-check window. Pricing, availability and security details move fast, so confirm those with the vendor.

ClaimSourceChecked
Codex executes coding tasks in sandboxes and submits pull requestsOpenAI Codex documentationJun 2026
Codex credits get the same Aug 21 GPT-5.6 Sol price cut as the API: input down 20% to $4, output down 33% to $20 per M tokens, through Nov 21· re-checkWinbuzzer; Business Standard ↗Aug 2026
OpenAI opened a public-beta Agents API (Sept 10, 2026) exposing Codex's own session orchestration, context compaction and crash-recovery harness to any developer, priced at pass-through token and tool costs with no added feeMarkTechPost; OpenAI ↗Sep 2026
Plugin4Shell, a zero-click RCE affecting plugin SHA-pinning across four AI coding agents, was disclosed Sept 17, 2026; OpenAI had already closed the hole in Codex version 0.146.0 before the public disclosureThe Hacker News; Help Net Security ↗Sep 2026

Intelligence on Codex

Securityhigh impactSep 17, 2026

Plugin4Shell: a zero-click RCE hits Claude Code, Codex, Copilot and Gemini CLI at once

Researchers disclosed Plugin4Shell on September 17, a flaw in how Claude Code, OpenAI Codex, GitHub Copilot and Gemini CLI verify plugin updates pinned to a specific Git commit hash. An attacker who controls the plugin's repository can create a branch named after that pinned SHA and set it as the default branch, so the agent pulls malicious code while the pin still looks honored. No click, no approval prompt, no reinstall. Plugins loaded this way inherit the developer's own permissions: local source, cloud credentials, SSH keys, sometimes production access. Anthropic fixed Claude Code in version 2.1.179 and OpenAI closed it in Codex 0.146.0, both before the public disclosure. Google called Gemini CLI deprecated and pointed users to Antigravity instead of patching it. Microsoft, notified back in June, still hadn't shipped a fix, published an advisory, or given a timeline as of September 20 — GitHub says its own branch-naming rules block the exploit on github.com, but Copilot also pulls from Bitbucket, GitLab and self-hosted marketplaces that carry no such protection.

Claude CodeCodexGitHub CopilotThe Hacker News; Help Net Security; The Register
Releasemedium impactSep 10, 2026

OpenAI puts the Codex agent harness behind a public Agents API

OpenAI opened its Agents API in public beta, exposing the same session orchestration, context compaction, sub-agent coordination and crash recovery that runs Codex to any developer through four primitives: agent, environment, session and events. There's no markup on top of it — customers pay only for model tokens, tool calls and whatever sandbox compute they choose, sourced from OpenAI or partners including Cloudflare, Vercel and Oracle. It's a direct pitch against LangChain, CrewAI and the rest of the orchestration layer: don't build the agent runtime, rent OpenAI's.

CodexMarkTechPost; OpenAI
Pricingmedium impactAug 21, 2026

OpenAI cuts GPT-5.6 Sol API pricing by up to a third, its second cut in a month

OpenAI dropped GPT-5.6 Sol's input price from $5 to $4 per million tokens and output from $30 to $20, a 20% and 33% cut respectively. The new rate covers the pay-as-you-go API, Codex credits and eligible ChatGPT Work plans through November 21; Plus, Pro and Business consumer subscriptions stay where they were. It's Sol's second markdown since the GPT-5.6 family launched in July. Google's Gemini 3.7 Flash shipped this week at roughly half the price of its predecessor, and Anthropic has been picking off enterprise accounts on reliability rather than price.

ChatGPTCodexWinbuzzer; Business Standard
Adoptionhigh impactAug 13, 2026

IBM joins OpenAI's Elite partner tier, wiring GPT-5.6 and Codex into IBM Consulting

IBM is bringing OpenAI's frontier models — GPT-5.6, Codex and ChatGPT Work — into IBM Consulting Advantage, the platform IBM uses to deliver consulting work to enterprise clients, with financial services, government, telecom and retail named as first targets. IBM becomes a member of OpenAI's Elite partner tier as part of the deal, and the two companies frame the collaboration around three areas: converting legacy workflows to AI-ready operations, modernizing application development, and cybersecurity/AI risk management. The frontier-model layer and the systems-integrator layer are consolidating. For anyone already inside an IBM Consulting engagement, those stop being two separate vendor decisions.

ChatGPTCodexIBM Newsroom announcement (newsroom.ibm.com); TechCrunch and Dataconomy coverage

Alternatives in Coding Assistants

Compare them →

GitHub / Microsoft

88.3Excellent

The most widely deployed AI pair programmer, native to GitHub.

Coding AssistantsSubscriptionMature
From $10 per month (Pro)

Anthropic

87.1Excellent

Terminal-first agentic coding with frontier Claude models.

Coding AssistantsSubscriptionEstablished
From $20 per month (via Claude Pro)

Anysphere

86.7Excellent

The AI-native code editor that made agentic coding mainstream.

Coding AssistantsFreemiumEstablished
From Free / $20 per month (Pro)