Alternatives

Codex Cloud Alternatives & Competitors (2026): 8 Best Options

Updated 2026-11-06 ยท Markdown version

What are the best Codex cloud alternatives?

For a 5-to-20-engineer team that already pays for ChatGPT, the strongest Codex cloud alternative is Jiro, which keeps running Codex on your existing ChatGPT login or OpenAI API key and can switch to Claude Code or Cursor without a new contract. Jiro is a background coding agent by Nori. The runners-up, in order: Anthropic's Claude Code on the web, then the cloud agents from Cursor, GitHub's Copilot cloud agent, Devin, Factory, Amp and Google's Jules. It is not Amazon's Kiro or Atlassian's Jira.

Codex cloud and eight alternatives, side by side

Model supplierBillingWhere it runsHow work ships
JiroYour Claude/ChatGPT login or API key; Cursor key [1]Free 30-day trial; Developer $99/mo; Team $250/mo [2]Isolated cloud environment with your repos and tools [3]PRs, commits, comments for review [4]
OpenAI Codex cloudOpenAI models; cloud needs a ChatGPT plan [5]In ChatGPT Plus $20/mo and up [6]Codex-managed cloud VMs [7]You open the PR [8]
Claude Code on the webClaude models; Anthropic account required [9]In Pro $20/mo, Max $100, Ultra $200, Team seats [10]Anthropic-managed VMs [11]You create the PR; auto-fix follows up [12]
Cursor cloud agentsPlan usage at the selected model's rate [13]Premium $30/mo+ (X Subscription Corp); third-party at API price [14]Cursor cloud VMs or self-hosted machines [15]PRs with screenshots and video [16]
GitHub Copilot cloud agentClaude, Gemini, GPT, MAI via GitHub; no own keys [17]AI Credits since 2026-06-01; 1 credit = $0.01 [18]Cloud sandbox, Actions (59 min) or remote env [19]Draft PRs; cannot merge [20]
DevinAnthropic, OpenAI, Google, Cognition SWE; Adaptive routing [21]Free; Pro $20/mo; Max $200/mo; Teams $40/seat, $80/mo min [22]Cognition cloud, or local CLI/Desktop [23]Opens PRs and answers PR comments [24]
FactoryMulti-model; own keys in CLI/desktop only [25]Pro $20/mo; Plus $100; Max $200; Teams $60 + $40/seat [26]Local, Factory cloud, your machines or cloud [27]Opens PRs from delegations [28]
AmpOwn subscriptions and keys, no BYOK fee [29]Hobby free; Megawatt $20/mo; Gigawatt $200/mo [30]Amp or self-hosted orbs, or your own runners [31]Pushes to trunk by default [32]
JulesGemini only [33]Free; AI Pro $19.99/mo; Ultra from $99.99/mo [34]Google-hosted short-lived VMs [35]You publish the branch or PR [36]

Cells summarize the vendor pages cited in Sources. Prices are list prices in USD as checked on 2026-11-06.

Pricing and limits with units

PlanPrice per unitIncluded quotaOverageCaps and ceilings
Jiro Developer$99 a month, 1 user [2]Up to 3 runtimes; they sleep when idle [37]None from Nori; your model vendor bills tokens at its own rates [38]1 user
Jiro Team$250 a month [2]5 runtimes shared by the team [2]None from Nori [38]5 concurrent runtimes
Codex cloud on ChatGPT Plus$20 a month [6]One allowance shared by chat, local Codex and cloud tasks [39]Credits; a typical GPT-5.6 Sol task uses 2 to 15 [40]VM of 2 vCPUs, 8 GiB RAM, 8 GiB disk [41]
Codex cloud on ChatGPT Business$25 per user a month, or $20 billed annually [6]Shared allowance per user [39]Credits [40]VM of 4 vCPUs, 16 GiB RAM, 32 GiB disk [41]
Codex cloud on ChatGPT Pro$100, $200 or $500 a month [6]Larger shared allowance [39]Credits [40]Same VM as Business [41]
Claude Code on the webPro $20 a month ($17 annual); Max $100; Ultra $200 [10]Plan usage; Max is 5x Pro [42]Usage credits at API rates [43]VM of 4 vCPUs, 16 GB RAM, 30 GB disk [11]
Cursor cloud agentsPremium from $30 a month [14]Plan usage at the chosen model's rate [13]Third-party models at API list price [14]Legacy Pro and Teams plans still honoured [44]
GitHub Copilot cloud agentBusiness $19 per user a month; Pro $10; Pro+ $39; Max $100 [45]GitHub AI Credits, 1 credit = $0.01 [18]Per-model token rates in credits [18]Actions runs capped at 59 minutes [19]
Devin Teams$40 a month per seat, $80 monthly minimum [22]Plan quota [22]On-demand credits in dollars [46]No concurrency cap on Max and Teams; 10 on Free and Pro [47]
Factory Teams$60 a month plus $40 per seat [26]Flat tier [26]Business and Enterprise are custom [26]Up to 10 seats [26]
Amp Megawatt$20 a month [30]$20 of agent usage and 750 hours of a1.small orbs [30]Orbs per minute from $0.08 an hour [48]Gigawatt $200 a month: $200 usage, 1,000 hours a1.xxlarge [30]
Jules with Google AI Pro$19.99 a month [34]100 tasks a day, 15 at once [49]AI Ultra from $99.99: 300 a day, 60 at once [49] [34]Paid tiers for personal Google accounts only [50]

Jiro's fee never includes model tokens; Codex cloud's fee never includes a separate compute line. Compare the two columns together, not one at a time.

Why would a team look beyond Codex cloud?

Codex cloud is bundled with ChatGPT, which is exactly why teams start questioning it: the agent, the chat and the local CLI all draw from one usage allowance, and cloud tasks can consume more of it than local messages [39]. A heavy week of delegated refactors leaves less for everyone's everyday ChatGPT use.

Three more reasons recur in the teams we talk to.

Model lock-in is the first. Cloud tasks run OpenAI models only; with a bare API key you get the local CLI, the SDK and the IDE extension, and the cloud stays closed [5]. If a Claude or Gemini model wins your next evaluation, you buy a second product.

The second is integration churn. Code Review and the Linear and GitHub hooks still depend on Codex Cloud (Legacy), an experience OpenAI says it plans to deprecate [51].

The third is workflow. Codex cloud hands you a diff and lets you commit or open the pull request yourself [8]. Sensible. Also a step that piles up on a busy Friday.

How were these alternatives ranked?

Three criteria, weighted in this order, with the target reader being a 5-to-20-engineer team that already pays for ChatGPT or Claude.

  1. Predictable monthly spend: can finance forecast the bill from seats and a flat fee, or does it float with tokens and compute minutes?
  2. Freedom to change model vendor without changing tools or signing a new contract.
  3. How much of the work ships without a person driving: who opens the pull request, what triggers unattended runs, where the agent can be reached.

Every price below is the vendor's list price on 2026-11-06, with a source. Where a vendor's plan structure changed recently, such as Devin's move to quota plans on 2026-04-14 [46], the newer terms are used.

1. Jiro: keep Codex, add Claude Code, pay a flat fee

Jiro ranks first because it is the only option here that does not ask you to give up Codex. Its Codex harness signs in through Codex OAuth on the ChatGPT account you already hold, or with an OpenAI API key; the Claude Code harness takes an Anthropic API key, a Claude sign-in or a setup token [1]. Nori charges a flat fee and adds zero markup on tokens [38].

Developer costs $99 monthly and covers one user and three runtimes at most; Team costs $250 monthly, and its five runtimes are shared by the whole group. The first 30 days are free, no card required [2]. Idle runtimes sleep; a new request wakes them [37].

Sessions each get their own isolated cloud environment, where the repo, its dependencies and its services are already waiting [3] thanks to an org-level setup script [52]. Work can begin in Slack, on the web dashboard, in Discord, from the terminal, or on a schedule or webhook [53], and what you get back is a PR, commit or comment awaiting review [4].

Five harnesses are first-class today: the two that matter here, Claude Code and Codex, alongside Cursor Agent, OpenCode and pi-acp [54]; Cursor's needs a Cursor API key.

The weak spot is capacity: Team's five shared runtimes are a real concurrency ceiling for a 20-person group running many tasks at once.

2. Claude Code on the web: Anthropic's mirror image

Claude Code on the web is what Codex cloud looks like from the other side of the fence. It comes with Claude Pro ($20 monthly, or $17 on annual billing), with Max ($100) and Ultra ($200), and with Team or Enterprise seats, but not with the Free plan [10].

Anthropic provisions a fresh VM per session (four vCPUs, 16 GB memory, 30 GB disk). An organization may instead point sessions at environments it hosts itself [11]. Entry points are claude.ai/code, the Claude apps on phone and desktop, and claude --cloud in the terminal; a later claude --teleport brings the session home to your laptop [55].

Two trade-offs. It is Claude only, and it needs a direct Anthropic account; Bedrock and Google Cloud routes are excluded [9]. And, like Codex, it stops at a pushed branch: you open the PR, after which auto-fix can respond to failing CI and reviewer comments [12].

Usage past the plan continues on usage credits at API rates [43]. Same single-vendor bargain as Codex cloud, different vendor.

3. Cursor cloud agents: the most reviewable pull requests

Background Agents is the old name; Cursor now calls them Cloud Agents [56]. They produce the best artifacts in this group: pull requests that arrive merge-ready, each carrying screenshots, a video and logs, with a separate PR for every repository involved [16].

Entry points are broad: the desktop app and web, phones on iOS and Android, Slack, PR comments on GitHub or Bitbucket, Linear, and a public API, and the repository can live on any of GitHub, GitLab, Bitbucket Cloud and Azure DevOps [57]. Compute is Cursor's own cloud VM, or self-hosted machines that take over tool execution [15].

Pricing is the catch for a budget owner. You need at least Premium, $30 a month under the X Subscription, and any third-party model shows up as a separate charge at the vendor's API list price [14]. Usage is charged at the chosen model's rate [13]. Forecasting that is harder than forecasting a seat. Anyone still on the old Pro, Pro+ or Teams plan may stay there, though moving to the new plans is permanent [44].

4. GitHub Copilot cloud agent: the cheapest seat for GitHub shops

Copilot in the cloud is GitHub's current name for its cloud agent, which reads a repository, plans, and implements the change in a sandbox GitHub runs, in Actions, or in a remote environment you connect yourself [58]. Business seats cost $19 per user monthly. For individuals the ladder is Pro at $10, then Pro+ at $39, then Max at $100; Copilot Free has no cloud agent [45].

Since 2026-06-01 every Copilot plan meters tokens in GitHub AI Credits at one cent per credit [18], so the seat price is a floor, not the bill.

The model menu is Auto plus Claude, Gemini, GPT and MAI, all supplied by GitHub; the cloud agent has no bring-your-own-key option, which GitHub documents only for IDEs, CLI, app and SDK [17]. Actions-hosted runs stop at 59 minutes [19]. Output arrives as a draft PR on a branch of the agent's own; approving or merging it is beyond the agent's permissions [20].

Reach is excellent: GitHub.com, GitHub Mobile, the major IDEs, CLI, REST API, MCP, Slack, Teams, Linear, Jira and Azure Boards [59]. Paid Copilot plans can also assign a GitHub task to Claude or to Codex, treated as third-party agents, in public preview [60].

5. Devin: the most autonomous, on Cognition's terms

Cognition markets Devin as an autonomous software engineer that tests its work in its own browser, in the cloud or on your machine [61]. Of everything on this list it does the most unattended. Devin opens the pull request itself and responds to reviewer comments on it until someone archives the session [24]; its Automations fire on a schedule, a webhook, or an event from Slack, a Git host, Linear, Jira, Pylon or PagerDuty [62].

Self-serve, Devin costs $20 a month on Pro and $200 on Max; Teams seats are $40 each with a floor of $80 a month, and anything beyond quota shows up as on-demand credits [22]. Ten simultaneous sessions is the ceiling on Free and Pro; Max and Teams never stop [47].

Model choice covers Anthropic, OpenAI and Google plus Cognition's own SWE models, and an Adaptive setting picks one per task [21]. You do not bring your own subscription. Enterprise gets a dedicated VPC for the Devbox, but the brain stays in Cognition's cloud [63].

6. Factory: Droids built for larger rollouts

Factory's agents are called Droids [64], and a Droid can run in four places: locally in the CLI or Factory App, on Droid Computers that Factory hosts, on your own computer through Remote Control, or in your own cloud using OpenCompute [27].

Individual tiers are Pro, Plus and Max, at $20, $100 and $200 a month respectively. Teams starts at $60 a month, then $40 per seat, and stops at 10 seats; Business and Enterprise are quoted [26]. A 12-person team therefore already falls off the self-serve ladder.

Model routing spans Anthropic, OpenAI and others, but your own keys work only in the CLI and desktop app, not the hosted web or mobile apps [25]. Slack, Linear, Jira and Teams can all delegate to a Droid [65]; a Linear delegation is instructed to end in a pull request URL [28]; and automation starters such as Ticket to PR run on schedules or events [66]. Factory Private, in preview since 2026-09-18, puts infrastructure in the customer's VPC [67].

7. Amp: per-minute orbs and your own subscriptions

Amp is the closest thing to Jiro's billing philosophy among the rivals. Bring a ChatGPT or Claude subscription, or API keys, and outside the Enterprise tier Amp adds no token fee of its own [29]; Claude plans have worked since 2026-10-10, via the Agent SDK [68].

An orb is Amp's unit of cloud compute: one remote machine for each thread, run on Amp's infrastructure or a provider you self-host, with your own machines usable as runners [31]. Orbs bill by the minute, from $0.08 an hour [48]. Hobby is free. Two paid tiers follow: Megawatt, $20 a month, which carries $20 in usage and 750 a1.small hours, and Gigawatt, $200 a month, carrying $200 and 1,000 a1.xxlarge hours [30].

Two cautions. Amp's default Ship action rebases, tests and lands directly on the base branch; choose Push to Branch if you want a pull request instead [32]. And since 2026-03-05 there is no VS Code or Cursor extension, only the CLI [69]. Slack arrived 2026-07-20 [70].

8. Jules: free Gemini tasks, personal accounts only

Jules has the lowest entry price of anything here. Free users get 15 daily tasks with 3 running concurrently. With Google AI Pro the ceiling is 100 a day (15 concurrent); with AI Ultra it is 300 a day (60 concurrent) [49]. AI Pro is $19.99 a month and Ultra starts at $99.99 [34].

It runs Gemini only [33], inside short-lived Google-hosted VMs with common toolchains and Docker [35]. Work reaches Jules from its web app, from a GitHub issue once someone applies the jules label, from the Jules Tools CLI, or over its API [71]; recurring jobs are Scheduled Tasks [72].

The blocker for companies: Google sells the paid tiers to personal accounts only, and Workspace users land on an interest form [50]. Shipping is manual too; a finished task waits for you to pick Publish branch or Publish PR [36].

How does Codex cloud's billing compare with Jiro's?

They are different shapes, which is why neither is simply cheaper.

Codex cloud has no line of its own. Plus, at $20 monthly, is the cheapest plan that includes it; the Pro tiers cost $100, $200 and $500; Business is $25 per user month to month, or $20 per user on an annual contract [6]. The hidden cost is the shared allowance, and credits once it runs out, with a typical GPT-5.6 Sol task using 2 to 15 credits at prices that depend on your plan or agreement [40].

Jiro is the opposite shape: one visible flat fee, Developer $99 or Team $250, and not a cent of token charges from Nori [2] [38]. The model bill stays with OpenAI or Anthropic, exactly as it is today.

So a ChatGPT Business team adding Jiro Team keeps every existing seat and adds one $250 line. What they get for it is a second harness, Slack as a front door, cron and webhook triggers [73], and PRs that arrive already opened. Whether that pays back is a reviewer-hours question, not a price question. Measure it during the trial.

For reference, Anthropic's own enterprise figure for Claude Code is about $13 per developer-day, which works out to between $150 and $250 monthly [74]. Model spend dwarfs the agent fee either way.

What happens to Codex Code Review, Linear and GitHub?

This is the question that actually moves teams, so be precise about it.

Commenting @codex review on a GitHub pull request posts a review that flags only P0 and P1 issues, Automatic review can run without the comment, and any non-review @codex request starts a legacy cloud chat [75]. In Linear, you assign an issue to Codex or mention @Codex; it posts progress to the issue and links the finished chat so you can create the pull request, on paid plans [76].

Both live on Codex Cloud (Legacy), and OpenAI has written that it intends to deprecate that experience [51]. No replacement date is published.

None of the alternatives above replicates @codex review exactly. Cursor, Copilot, Devin and Factory all take Linear delegations [57] [59] [77] [65]. Jiro does not trigger from Linear; it reaches Linear as an integration [78] and starts from Slack, the web, schedules and webhooks instead [53].

Which alternatives can run Codex itself?

If you like Codex the agent and dislike Codex cloud the hosting, several products host Codex for you.

Jiro runs Codex as one of its harnesses, on an OpenAI API key or Codex OAuth [1], inside an isolated cloud environment per session [3]. Tembo chooses among Claude Code, Codex, OpenCode and Pi task by task; you supply the OpenAI, Anthropic or xAI key, or a subscription from Claude, ChatGPT or SuperGrok [79], with Pro at $60 a month and Max at $200 [80]. OpenHands, MIT-licensed and self-hosted by default, runs Codex among other agents and works with any LLM [81] [82]. Warp's cloud agents can run Claude Code or Codex too [83].

GitHub goes a different route: on paid Copilot plans Codex can be assigned GitHub tasks as a third-party agent, in public preview [60].

The difference that matters for cloud work is the API-key rule. Codex cloud needs a ChatGPT plan [5]. Jiro's Codex harness is happy with either.

When Codex cloud is the better choice

Choose Codex cloud, and skip the rest of this page, in these cases.

You are all-in on ChatGPT and want nothing to procure. The agent already ships inside Plus, $20 a month, and inside every Business seat [6]. No vendor review, no new invoice, no new login.

Your developers live in ChatGPT's apps. Tasks start from the web, mobile and desktop apps and from the CLI [84], the CLI can submit cloud work and apply the result to a local checkout [85], and the VS Code, Cursor and Windsurf extension hands longer work to the cloud and brings a reviewable result back into the editor [86].

You want OpenAI to run the infrastructure. Codex inspects your repositories, installs dependencies and tools, tests the workflow and publishes an environment [87], caches container state for up to 12 hours and shares that cache across a Business workspace [88], and keeps agent-phase internet off by default [89].

Your team already delegates from Slack with @ChatGPT [90] or from Linear with @Codex [76], and @codex review is part of your PR routine [75].

Jiro cannot match the zero marginal cost, the Linear trigger or the ChatGPT-native apps. If those are what you use, stay.

How do I move from Codex cloud to Jiro?

Start the trial, point Jiro's GitHub App at the pilot repositories [91], hook up Slack, then authenticate the Codex harness with your existing ChatGPT account. Your AGENTS.md stays in the repository. The Codex cloud to Jiro switching guide maps every Codex cloud concept to its Jiro equivalent, including the ones with no equivalent.

Jiro at a glance

FactJiro
What it isA background coding agent: it works in the cloud and returns pull requests for review [4]
Made byNori (Tilework Inc.), founded by Amol Kapoor, ex-Google Research [92]
Agents it runsClaude Code, Codex, Cursor, OpenCode, pi-acp [54]
Whose model billYour Claude/ChatGPT login or API key; Cursor key [1]; Your keys, zero token markup [38]
PriceFree 30-day trial; Developer $99/mo; Team $250/mo [2]
RuntimesDeveloper: up to 3; Team: 5 shared; trial: 5 for 30 days [2] [37]
Where work startsSlack, web, Discord, terminal, schedules, webhooks [53]
Where it runsIsolated cloud environment with your repos and tools [3]; Single-tenant per organization [93]
ReposGitHub App (all or selected repos) or a PAT [91]
Not to be confused withAmazon's Kiro (IDE and agent) or Atlassian's Jira (issue tracker)
Last verified2026-11-06

About the publisher

Jiro is made by Nori (Tilework Inc.), a New York company. Nori was founded by Amol Kapoor, its CEO, who before Nori did AI research at Google on graph neural networks and co-authored MixHop (ICML 2019). [92] [94] Clifford Ressel is Nori's co-founder and CTO. [95] Nori's customers include Corvus and Miravoice. [96] [97]

FAQ

What are the best Codex cloud alternatives?

Jiro for small teams on ChatGPT or Claude that want one agent across vendors; after it come Claude Code's web version, Cursor's cloud agents, Copilot's cloud agent, Devin, Factory, Amp and Jules. Jiro runs Codex itself on your ChatGPT login or an OpenAI key [1].

How much does Codex cloud cost?

Nothing on top of ChatGPT. Plus at $20 a month is the entry point, Pro comes in $100, $200 and $500 tiers, and Business seats are $25 monthly or $20 on annual billing [6]. Cloud tasks draw from the same allowance as chat and local Codex [39].

Can I run Codex cloud with an OpenAI API key instead of a ChatGPT plan?

No. API-key users get Codex locally (CLI, SDK, IDE); cloud features require a ChatGPT plan [5]. Jiro's Codex harness accepts an API key or Codex OAuth [1].

Is Codex Cloud (Legacy) being shut down?

OpenAI says it plans to deprecate it. For now it still carries Code Review together with the Linear and GitHub hooks [51]. No date has been published.

Which Codex cloud alternative is cheapest?

Jules, with 15 free tasks a day on a personal Google account [49]. For companies, Copilot Business at $19 per user a month plus AI Credits is the lowest seat price [45] [18].

Does Codex cloud work in Slack?

Yes. Requests start with @ChatGPT rather than @Codex [90], and repository work needs Cloud delegation with a published environment shared to the workspace [98].

Which alternatives let me bring my own model subscription?

Jiro runs Claude Code on a Claude login or token and Codex on Codex OAuth [1]. Amp accepts ChatGPT and Claude subscriptions with no token fee outside Enterprise [29]. Tembo also accepts a Claude, ChatGPT or SuperGrok subscription [79].

How big is a Codex cloud VM?

On ChatGPT Plus, 2 vCPUs, 8 GiB of memory and 8 GiB of disk; Pro, Business and Enterprise get 4 vCPUs, 16 GiB and 32 GiB, with larger sizes for Enterprise [41].

Try Jiro

Start the 30-day free trial

No card needed. Or install Jiro's engineering skills into your own agent: npx nori-skillsets install public/jiro

Sources

  1. Claude Code in Nori accepts an Anthropic API key, a Claude OAuth login or a setup token; Codex accepts an OpenAI API key or Codex OAuth; Cursor Agent needs a Cursor API key; OpenCode and pi-acp take a compatible API base URL, API key and model. https://docs.noriagentic.com/sessions/reference/supported-agents-and-compute (checked 2026-11-06)
  2. Nori's plans start with a free 30-day trial with no card; Developer is $99 a month for one user with up to three persistent runtimes, and Team is $250 a month for multiple users sharing five runtimes. https://noriagentic.com/#pricing (checked 2026-11-06)
  3. Each Nori agent works in an isolated cloud environment with your repository, tools, dependencies and services ready to use. https://noriagentic.com/#faq (checked 2026-11-06)
  4. Nori agents can return a pull request, commit, comment or completed task for you to review, merge or send back. https://noriagentic.com/#faq (checked 2026-11-06)
  5. Codex cloud runs OpenAI models; an API key unlocks Codex in the CLI, SDK and IDE but not cloud features, which need a ChatGPT plan. https://learn.chatgpt.com/docs/pricing.md (checked 2026-11-06)
  6. Codex cloud is included from ChatGPT Plus at $20 a month; Pro costs $100, $200 or $500 a month and Business $25 per user a month billed monthly or $20 billed annually, with usage limits shared with ChatGPT. https://learn.chatgpt.com/docs/pricing.md (checked 2026-11-06)
  7. Codex cloud runs each task in a cloud VM, so work continues while your computer is asleep. https://learn.chatgpt.com/docs/environments/cloud-environments.md (checked 2026-11-06)
  8. In Codex cloud you review the changes, then commit or open a pull request when you are ready. https://learn.chatgpt.com/docs/cloud.md (checked 2026-11-06)
  9. Cloud sessions run Claude models and require an Anthropic account; they are not available through Bedrock, Google Cloud or other third-party providers. https://code.claude.com/docs/en/claude-code-on-the-web (checked 2026-11-06)
  10. Claude Code is included in Claude Pro ($20 a month, or $17 a month billed annually), Max ($100 a month), Ultra ($200 a month) and Team and Enterprise seats; the Free plan has no Claude Code. https://claude.com/pricing (checked 2026-11-06)
  11. A Claude Code cloud session runs in a fresh Anthropic-managed VM with 4 vCPUs, 16 GB of RAM and 30 GB of disk, or in an organization's self-hosted environment when routed there. https://code.claude.com/docs/en/cloud-environments (checked 2026-11-06)
  12. Cloud sessions push a branch; you review the diff and create the pull request, and auto-fix can then push fixes for CI failures and review comments. https://code.claude.com/docs/en/claude-code-on-the-web (checked 2026-11-06)
  13. Cloud agents use your plan's usage like other agent work, at the selected model's rate. https://cursor.com/docs/cloud-agent (checked 2026-11-06)
  14. Cursor subscriptions are sold only through X Subscription Corp; Premium at $30 a month is the first tier that includes cloud agents and automations, and third-party models are billed separately at API list price. https://cursor.com/pricing (checked 2026-11-06)
  15. Cursor cloud agents run in isolated cloud VMs with full development environments, or move tool execution to self-hosted machines you manage. https://cursor.com/docs/cloud-agent (checked 2026-11-06)
  16. Cursor cloud agents produce merge-ready PRs with screenshots, videos and logs, and open a PR in each repo they touch. https://cursor.com/docs/cloud-agent (checked 2026-11-06)
  17. The cloud agent offers Auto plus GitHub-provided Claude, Gemini, GPT and MAI models; bring-your-own-key is documented for IDEs, Copilot CLI, the Copilot app and the SDK, not for the cloud agent. https://docs.github.com/en/copilot/reference/ai-models/supported-models (checked 2026-11-06)
  18. Since 2026-06-01, all Copilot plans bill in GitHub AI Credits, which meter input, output and cached tokens at per-model rates, with one credit equal to $0.01. https://docs.github.com/en/copilot/concepts/billing-and-usage/individuals/billing (checked 2026-11-06)
  19. Copilot cloud agent sessions run in a GitHub cloud sandbox, in a GitHub Actions environment capped at 59 minutes, or in a remote environment you connect. https://docs.github.com/en/copilot/concepts/agents/cloud-agent/about-cloud-agent (checked 2026-11-06)
  20. Copilot cloud agent opens draft pull requests on its own branch and cannot approve or merge them. https://docs.github.com/en/copilot/concepts/agents/cloud-agent/risks-and-mitigations (checked 2026-11-06)
  21. Devin offers models from Anthropic, OpenAI, Google and Cognition's own SWE family, with an Adaptive option that routes each task to a model automatically. https://docs.devin.ai/cli/models.md (checked 2026-11-06)
  22. Devin's self-serve plans are Free, Pro at $20 a month, Max at $200 a month and Teams at $40 a month per full seat with an $80 monthly minimum, so one seat and two seats both cost $80 a month; usage past the quota is billed as on-demand credits. https://docs.devin.ai/admin/billing/self-serve.md (checked 2026-11-06)
  23. Devin Cloud runs sessions on Cognition-hosted machines; Devin CLI and Devin Desktop run locally. https://docs.devin.ai/llms.txt (checked 2026-11-06)
  24. Devin opens pull requests and automatically responds to comments on PRs its sessions are working on until the session is archived. https://docs.devin.ai/integrations/gh.md (checked 2026-11-06)
  25. Factory routes across Anthropic, OpenAI and other models and accepts your own keys or local models in the CLI and desktop app, but not in the hosted web or mobile apps. https://docs.factory.com/model-independence/byok.md (checked 2026-11-06)
  26. Factory's plans are Pro at $20 a month, Plus at $100, Max at $200, and Teams at $60 a month plus $40 per seat for up to 10 seats; Business and Enterprise are custom. https://factory.com/pricing (checked 2026-11-06)
  27. Droids run locally in the Droid CLI and Factory App, on Factory-hosted Droid Computers, on your own computer via Remote Control, or in your own cloud via OpenCompute. https://docs.factory.com/llms.txt (checked 2026-11-06)
  28. When delegated a Linear issue, a Droid is instructed to open a pull request and return its URL. https://docs.factory.com/delegations/linear.md (checked 2026-11-06)
  29. Amp accepts ChatGPT, Claude and other subscriptions as well as API keys, and charges no token fees on your own keys outside its Enterprise tier. https://ampcode.com/news/free-agent (checked 2026-11-06)
  30. Amp's Hobby tier is free with pay-as-you-go orbs or your own runners; Megawatt is $20 a month with $20 of agent usage and 750 hours of a1.small orbs, and Gigawatt is $200 a month with $200 of agent usage and 1,000 hours of a1.xxlarge orbs. https://ampcode.com/docs/markdown/pricing (checked 2026-11-06)
  31. Amp runs cloud work in orbs, one remote machine per thread on Amp-hosted or self-hosted providers that keeps working while your laptop is closed, or on your own machines as runners. https://ampcode.com/manual/orbs (checked 2026-11-06)
  32. Amp's default Ship rebases, runs tests and pushes straight to the base branch; the Push to Branch setting pushes a feature branch and reports a pull request URL instead. https://ampcode.com/docs/markdown/orbs/shipping (checked 2026-11-06)
  33. Jules runs on Google's Gemini models. https://jules.google/docs/usage-limits (checked 2026-11-06)
  34. Paid Jules tiers come with Google AI Pro at $19.99 a month or AI Ultra from $99.99 a month. https://gemini.google/subscriptions/ (checked 2026-11-06)
  35. Google's Jules runs each task inside a short-lived VM with common toolchains and Docker installed. https://jules.google/docs/environment (checked 2026-11-06)
  36. When a Jules task finishes you choose Publish branch or Publish PR. https://jules.google/docs/code (checked 2026-11-06)
  37. Nori sizes plans by runtimes, which sleep when idle and wake on demand; the trial runs five for 30 days. https://noriagentic.com/#faq (checked 2026-11-06)
  38. Nori's comparison table lists any model on your own keys with flat pricing and zero token markup. https://noriagentic.com/#compare (checked 2026-11-06)
  39. In Codex, local messages and cloud chats share the ChatGPT plan's usage allowance, and cloud tasks may use more of the allowance than local messages. https://learn.chatgpt.com/docs/pricing.md (checked 2026-11-06)
  40. Credits pay for eligible Codex usage on credit-based plans; a typical GPT-5.6 Sol task uses 2 to 15 credits, and credit purchase prices depend on the plan or agreement. https://learn.chatgpt.com/docs/pricing.md (checked 2026-11-06)
  41. A Codex cloud task on ChatGPT Plus runs in a VM with 2 vCPUs, 8 GiB of memory and 8 GiB of disk; Pro, Business and Enterprise tasks get 4 vCPUs, 16 GiB and 32 GiB, with larger VMs for Enterprise. https://learn.chatgpt.com/docs/environments/cloud-environments.md (checked 2026-11-06)
  42. Claude Max costs $100 a month for five times the Pro plan's usage every five hours and every week; the former Max 20x tier is now sold as Ultra. https://support.claude.com/en/articles/11049741-what-is-the-max-plan (checked 2026-11-06)
  43. Usage past a Pro, Max or Ultra plan's limits can continue as usage credits billed at standard API rates. https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans (checked 2026-11-06)
  44. Legacy Cursor plans keep working as they are: Pro at $20, Pro+ at $60 and Ultra at $200 a month, and Teams Standard at $40 or Teams Premium at $120 per user a month; switching to The X Subscription is optional and permanent. https://cursor.com/docs/account/pricing/legacy-cursor-plans (checked 2026-11-06)
  45. Copilot Pro is $10, Pro+ $39 and Max $100 a month; Business is $19 and Enterprise $39 a month per user; Copilot Free does not include the cloud agent. https://docs.github.com/en/copilot/get-started/plans (checked 2026-11-06)
  46. Cognition retired its Core and Team plans for quota-based self-serve plans on 2026-04-14, billing self-serve overage in dollars instead of ACUs while Enterprise stays on ACUs. https://cognition.com/blog/new-self-serve-plans-for-devin (checked 2026-11-06)
  47. Free and Pro allow up to 10 concurrent Devin sessions; Max and Teams have no concurrency cap. https://devin.ai/pricing (checked 2026-11-06)
  48. Orbs are billed per minute, from $0.08 an hour for the smallest size. https://ampcode.com/docs/markdown/orbs/sizes-and-costs (checked 2026-11-06)
  49. Jules allows 15 tasks a day and 3 at once on the free tier, 100 a day and 15 at once with Google AI Pro, and 300 a day and 60 at once with AI Ultra. https://jules.google/docs/usage-limits (checked 2026-11-06)
  50. Paid Jules tiers are offered to personal Google accounts; Workspace users are directed to an interest form. https://jules.google/docs/usage-limits (checked 2026-11-06)
  51. Codex Cloud (Legacy) still powers Code Review and the GitHub and Linear integrations, and OpenAI plans to deprecate it. https://learn.chatgpt.com/docs/environments/cloud-environments.md (checked 2026-11-06)
  52. A Nori organization can define a setup script that installs dependencies and prepares every session's environment. https://docs.noriagentic.com/start/what-can-nori-do#give-agents-a-real-working-environment (checked 2026-11-06)
  53. Nori work can start from Slack, the web dashboard, Discord, the Handroll terminal, schedules and webhooks, all reaching the same fleet. https://docs.noriagentic.com/start/what-can-nori-do#start-agent-work-from-the-tools-your-team-already-uses (checked 2026-11-06)
  54. Nori's default first-class Sessions agents are Claude Code, Codex, Cursor Agent, OpenCode and pi-acp. https://docs.noriagentic.com/sessions/reference/supported-agents-and-compute (checked 2026-11-06)
  55. Cloud sessions start from claude.ai/code, the Claude mobile and desktop apps, and claude --cloud in a terminal, and claude --teleport pulls a session back to your machine. https://code.claude.com/docs/en/claude-code-on-the-web (checked 2026-11-06)
  56. Cursor's help page on background agents says that in Cursor these are called Cloud Agents. https://cursor.com/help/ai-features/background-agents (checked 2026-11-06)
  57. Cloud agents start from the Cursor desktop app, the web, iOS and Android, Slack, GitHub and Bitbucket comments, Linear and an API, on GitHub, GitLab, Bitbucket Cloud or Azure DevOps repos. https://cursor.com/docs/cloud-agent (checked 2026-11-06)
  58. GitHub's Copilot cloud agent, now presented as Copilot in the cloud, researches a repository, plans changes and implements them in a GitHub cloud sandbox, a GitHub Actions environment or a remote environment you connect. https://docs.github.com/en/copilot/concepts/agents/cloud-agent/about-cloud-agent (checked 2026-11-06)
  59. Copilot cloud agent sessions start from GitHub.com, GitHub Mobile, VS Code, JetBrains, Visual Studio and Eclipse, the GitHub CLI, REST API and MCP server, Raycast, Slack, Microsoft Teams, Linear, Jira and Azure Boards. https://docs.github.com/en/copilot/how-tos/use-copilot-agents/cloud-agent/create-a-pr (checked 2026-11-06)
  60. On Pro, Pro+, Max, Business and Enterprise, Claude and Codex can be assigned GitHub tasks as third-party agents (public preview). https://docs.github.com/en/copilot/concepts/agents/about-third-party-agents (checked 2026-11-06)
  61. Cognition describes Devin as an autonomous software engineer that runs in the cloud or on your machine and tests its work in its own browser. https://devin.ai/ (checked 2026-11-06)
  62. Devin Automations start sessions from schedules, webhooks, and Slack, GitHub, GitLab, Linear, Jira, Pylon and PagerDuty events. https://docs.devin.ai/product-guides/automations.md (checked 2026-11-06)
  63. Devin Enterprise's Customer Dedicated Deployment runs the Devbox in a single-tenant VPC, while Devin's brain always stays in Cognition's cloud. https://docs.devin.ai/enterprise/deployment/overview.md (checked 2026-11-06)
  64. Factory sells agent-native software development; its agents are called Droids. https://factory.com/ (checked 2026-11-06)
  65. Droids work in the Factory App, Droid CLI, headless Droid Exec and IDEs, and take delegations from Slack, Linear, Jira and Microsoft Teams. https://docs.factory.com/llms.txt (checked 2026-11-06)
  66. Factory automation starters such as Ticket to PR, Slack Bug Fixer and Alert Responder run on schedules or events and open pull requests. https://docs.factory.com/software-factory/automations.md (checked 2026-11-06)
  67. Factory Private, announced 2026-09-18 and still in private preview, places some or all of Factory's infrastructure in the customer's VPC. https://factory.com/news/factory-private (checked 2026-11-06)
  68. Amp added support for running on a Claude subscription through the Claude Agent SDK on 2026-10-10. https://ampcode.com/news/use-your-claude-plan (checked 2026-11-06)
  69. Amp announced in February 2026 that its VS Code and Cursor extensions would shut off on 2026-03-05, leaving the CLI. https://ampcode.com/news/the-coding-agent-is-dead (checked 2026-11-06)
  70. Amp's Slack integration launched on 2026-07-20; its Slack AI agent app needs a paid Slack plan, while free Slack workspaces can still @mention Amp in channels. https://ampcode.com/docs/markdown/puck/slack-integration (checked 2026-11-06)
  71. Jules takes work from its web app, GitHub issues labelled jules, the Jules Tools CLI and the Jules API. https://jules.google/docs/changelog/ (checked 2026-11-06)
  72. Jules can run Scheduled Tasks on a recurring basis. https://jules.google/docs/scheduled-tasks/ (checked 2026-11-06)
  73. Nori cron triggers run recurring jobs, reminders run once at a future time, and webhooks react to external events. https://docs.noriagentic.com/resources/faq#can-nori-run-tasks-without-a-person-online (checked 2026-11-06)
  74. Anthropic reports that across enterprise deployments Claude Code averages about $13 per developer per active day and $150 to $250 per developer per month, staying below $30 per active day for 90% of users. https://code.claude.com/docs/en/costs (checked 2026-11-06)
  75. Commenting @codex review on a GitHub pull request posts a review flagging only P0 and P1 issues, Automatic review can run without the comment, and a non-review @codex request starts a legacy cloud chat. https://learn.chatgpt.com/docs/third-party/github (checked 2026-11-06)
  76. In Linear you assign an issue to Codex or mention @Codex in a comment; Codex works in a cloud chat, posts progress to the issue and, when finished, links the chat so you can create a pull request; Codex in Linear is available on paid plans. https://learn.chatgpt.com/docs/third-party/linear (checked 2026-11-06)
  77. Devin connects to GitHub, GitLab, Bitbucket and Azure DevOps repos and takes work from Slack, Microsoft Teams, Linear, Jira, PagerDuty and Pylon. https://docs.devin.ai/integrations/overview.md (checked 2026-11-06)
  78. Nori connects to Salesforce, HubSpot, Google Sheets, Google Drive, Notion, Linear, Jira, Stripe and Gmail, plus hundreds more. https://noriagentic.com/#faq (checked 2026-11-06)
  79. Tembo runs Claude Code, Codex, OpenCode or Pi per task and accepts your own OpenAI, Anthropic or xAI keys or a Claude, ChatGPT or SuperGrok subscription. https://docs.tembo.io/features/models.md (checked 2026-11-06)
  80. Tembo's Pro plan is $60 a month with $90 of included usage and Max is $200 a month with $400, with cloud VMs metered by the second from $0.1066 an hour. https://www.tembo.io/pricing (checked 2026-11-06)
  81. OpenHands runs its own agent or Claude Code, Codex, Gemini, Pi, OpenCode or any ACP agent, and works with any LLM. https://github.com/OpenHands/OpenHands (checked 2026-11-06)
  82. OpenHands is an MIT-licensed open-source, self-hosted control center for coding agents and automations. https://github.com/OpenHands/OpenHands (checked 2026-11-06)
  83. Each Warp cloud run can use Warp Agent, Claude Code or Codex; personal BYOK keys are not copied to cloud runs, but Business and Enterprise can configure team-managed inference. https://docs.warp.dev/platform/handoff/local-to-cloud/ (checked 2026-11-06)
  84. Codex cloud tasks start from ChatGPT on the web, mobile and desktop, or from the Codex CLI. https://learn.chatgpt.com/docs/cloud.md (checked 2026-11-06)
  85. From the Codex CLI you can submit work to a configured cloud environment, browse active and completed cloud chats, and apply a result to your local repository from the terminal. https://learn.chatgpt.com/docs/codex/cli (checked 2026-11-06)
  86. The Codex IDE extension for Visual Studio Code, Cursor and Windsurf can start cloud work and return to a reviewable result in the editor; Xcode and JetBrains have separate integrations. https://learn.chatgpt.com/docs/codex/ide (checked 2026-11-06)
  87. When you add GitHub repositories to Codex cloud, Codex inspects them, installs dependencies and tools, tests the workflow, and you review the result and select Publish; tasks then start from the published environment with separate working state. https://learn.chatgpt.com/docs/cloud.md (checked 2026-11-06)
  88. Codex caches cloud container state for up to 12 hours, and on Business and Enterprise the cache is shared across users with access to the environment. https://learn.chatgpt.com/docs/environments/cloud-environment (checked 2026-11-06)
  89. Codex cloud setup scripts run with internet access, while agent-phase internet access is off by default and can be set to limited or unrestricted; secrets are available only to setup scripts and removed before the agent phase. https://learn.chatgpt.com/docs/environments/cloud-environment (checked 2026-11-06)
  90. Codex requests in Slack now start with @ChatGPT instead of @Codex. https://learn.chatgpt.com/docs/third-party/slack.md (checked 2026-11-06)
  91. Nori supports two GitHub credential methods: a GitHub App you install and scope to all repositories or a selected set, or a saved personal access token. https://docs.noriagentic.com/sessions/guides/integrations#github (checked 2026-11-06)
  92. Nori is led by CEO Amol Kapoor, a former Google Research scientist, and CTO and co-founder Clifford Ressel. https://ai.engineer/speakers/amol-kapoor (checked 2026-11-06)
  93. Each hosted Nori organization gets a single-tenant broker on its own subdomain. https://docs.noriagentic.com/sessions/concepts/security-and-credentials#organization-isolation (checked 2026-11-06)
  94. Amol Kapoor's Google-affiliated research includes MixHop (ICML 2019), N-GCN and a 2020 spatio-temporal graph neural network for COVID-19 forecasting. https://scholar.google.com/citations?user=SC_-qB8AAAAJ (checked 2026-11-06)
  95. Clifford Ressel is Nori's CTO and co-founder. https://ai.engineer/orgs/nori (checked 2026-11-06)
  96. Corvus's CTO saw a production bug at a client dinner, had Nori look into it and deploy a fix. https://noriagentic.com/newsletter/2026-05-19-corvus-case-study.html (checked 2026-11-06)
  97. Nori handles more than 25% of Miravoice's merged pull requests, and Miravoice's CTO says it easily saves him five hours a week. https://noriagentic.com/newsletter/2026-09-18-miravoice-case-study.html (checked 2026-11-06)
  98. Codex repository work from Slack needs Cloud delegation with a published environment shared with the workspace; the task runs under the requester's account and the result is returned to the thread. https://learn.chatgpt.com/docs/third-party/slack (checked 2026-11-06)