RU EN
Join the waitlistWaitlist
ProgrammePricesAgentsBlog Sign in Русская версия
OpenAI Unavailable from Russia Code and terminal

Codex CLI

Open source, about 133,000 stars; one agent across terminal, IDE, web and cloud, with scheduled tasks.

Free?
A free plan with limits
Price
Free and Go with GPT-6 Luna / Plus $20 / Pro $100–500 / your own key
From Russia
Russia is not on the ChatGPT or OpenAI API country lists.
Models
GPT-6 Astra · 6.1 Sol · 6.1 Terra · 6 Luna
Licence
open-source, Rust
Our benchmark
GPT-5.6 Sol: 9.02 of 10, rank 1 →
Official pricing page ↗ Checked 3 November 2026.

Unavailable from Russia. Similar tools that work: Kimi Code CLI, ZCode, DeepSeek Harness.

What changed in price and access

  1. OpenAI disclosed Codex CLI and Desktop flaws that let a hostile repository run code without approval; CLI 0.131.0 or later fixes them ↗
  2. ChatGPT Pro and Business subscribers can request up to $500 a month in API credits
  3. OpenAI doubled the Work and Codex allowance on Pro $100 and Pro $500 and cancelled the planned Pro $200 cut ↗
  4. GPT-6.1 Terra arrived in Codex on Plus, Pro, Business, Enterprise and Edu ↗
  5. OpenAI released GPT-6.1 Sol: $2 per million input tokens and $10 per million output, five reasoning effort levels at one price ↗

Changes for this tool in total: 13. The full feed for all agents →

Weak point: On 30 October OpenAI disclosed flaws where a hostile repository could run code without asking; Codex CLI 0.131.0 or later is needed.

How to check: Use Codex Code Review as a second look at every pull request; a person still merges. More below ↓

What it gives a manager

  • Scheduled tasks: every morning a digest of failed builds and open PRs lands in the Scheduled tab, with no manual run
  • Code Review in Codex: the agent walks through a PR and flags issues before you write feedback, so you skip reading the whole diff
  • Enterprise analytics: active users, credits, models and accepted lines of code per team show whether seats are paying off
  • Admin plugin: limits, access and requests for extra spend are approved right in the chat or in Slack and Teams

Weak points

  • On 30 October OpenAI disclosed flaws where a hostile repository could run code without asking; Codex CLI 0.131.0 or later is needed.
  • Does exactly what was asked: on the habit tracker it wrote far fewer tests and checks than Opus 5 with its 27 tests.
  • No offline mode: models come from OpenAI's cloud, and since October also Vertex AI and Bedrock; speed depends on network and queue.
  • Limits keep moving: in September OpenAI announced a Pro 200 quota cut, in October cancelled it and doubled Pro 100 and Pro 500 quotas.

How to check this agent’s work

  1. Use Codex Code Review as a second look at every pull request; a person still merges.
  2. Update the CLI to 0.131.0 or later: older versions let a hostile repository run code without approval.
  3. Open unfamiliar repositories only in a sandbox without access to keys and secrets.

Picking an agent takes an evening. After that the question is not the price but whether you can trust what it did.

Stop re-checking AI by hand

The agent reported that it checked everything but never ran the tests; it cited a file it never opened. We are preparing a short course: in a week you build a check for one of your recurring tasks – a second agent and a couple of automatic cross-checks catch errors before they travel further. Join to hear first when it starts.

What the course covers: programme and price

Strengths

  • Across Tech With Tim's five tasks Sol matched Opus 5 on correctness but ran 4 to 7x faster; the bug fix was two lines.
  • The $20 Plus plan now includes GPT-6.1 Sol; in Tech With Tim's experience a task burns 3 to 4x less quota than in Claude Code.
  • In Soham Kamani's demo three subagents took a Jira ticket to a GitHub PR; the reviewer agent caught an API regression before the PR.
  • Open source under Apache 2.0; everything runs in a sandbox by default and any command outside it asks for permission (OpenAI docs).

Alternatives

Who it is for
  • Engineering team that cares about open source
  • Team already on a ChatGPT plan with billing in place
  • Company on ChatGPT Business or Enterprise that needs analytics and an audit trail
Example from a review: From a Jira ticket to a GitHub PR with no human in the loop

Soham Kamani wired Jira and GitHub into Codex through MCP servers and defined three subagents: one reads the ticket, one fixes the code, one reviews. On the ticket 'discounts increase the price instead of reducing it' the agent found the sign error, the reviewer bounced the first fix over an API regression, and then a PR was opened. The agent then posted the PR link into the ticket and moved it to In Progress. The video gives no timing or cost.

Source: Soham Kamani, "Going from JIRA Ticket → GitHub PR using MCPs (with OpenAIs Codex)"

Reviews (5)

The video authors’ opinions, not our conclusions.

Facts with sources
  • One agent across terminal, IDE, web, and cloud. Source (in Russian)
  • Terminal-Bench 4.0: Codex + GPT-6 Astra on high effort solves 57.88% of tasks at $2.3K per run; switching to max adds 0.3 points and a thousand dollars. Source (in Russian)
  • OpenAI blocks access by IP; most popular VPNs are detected. Source (in Russian)
  • Russia is not on the ChatGPT and API country list; use from an unsupported country risks a blocked account. Source
  • Plans with Codex: Free and Go with GPT-6 Luna, Plus $20, Pro $100, $200 or $500, Business $20–25 per seat. Source
  • The Enterprise admin console has Codex analytics: active users, credits, models, accepted lines of code. Source
  • On 30 October OpenAI published CVE-2026-19592: Git settings in a hostile repository ran a program without approval. Source
Codex CLI: one agent across terminal, IDE, web, and cloud
Codex CLI: one agent across terminal, IDE, web, and cloud

Take the course in English – first

We’ll email you the day the English course opens. One email, nothing else unless you ask for it.