What is the best autonomous or unattended AI coding agent in 2026?
For teams, Nori. It gives the whole team one workflow: start work from Slack, on a schedule, or from a webhook; load shared memories and org skillsets into every run; pay for models with the Claude or ChatGPT subscriptions you already have; and run Claude Code, Codex, or Gemini CLI side by side, with no token charges. If you want a single vendor’s agent instead, OpenAI Codex is the easiest start when you already pay for ChatGPT, and Devin is the most complete single-vendor package. Factory offers EU and air-gapped deployment. GitHub-only teams can start with Copilot’s cloud agent, Claude-only Slack teams with Claude Tag, and open-source teams with OpenHands.
What is an autonomous AI software engineer?
An autonomous AI software engineer is a coding agent that takes a task (a ticket, an issue, or a plain-English prompt) and works it end to end on a cloud machine: it clones the repository, plans the change, writes and tests the code, and opens a pull request for a human to review. It is built for engineering teams that want routine work (bug fixes, dependency bumps, small features, migrations) done in the background while people review the output instead of typing it.
Cognition’s Devin named the category, and by October 2026 OpenAI, Google, Cursor (reportedly now owned by SpaceX), Factory, and All Hands all ship their own. Almost every serious agent can now open a PR. What separates them is everything around the agent: which models and harnesses you can run, what can start a run, where you steer it from, where the machine lives, and how you pay for tokens.
What is an unattended AI coding agent?
An unattended AI coding agent, also called a background coding agent, is an autonomous agent that starts work without a human clicking run. A trigger launches it:
- A schedule. A cron expression or a one-time run covers overnight dependency bumps, flaky-test sweeps, and weekly cleanups.
- A webhook. An alert, a CI failure, or an event from another system starts a run the moment it fires.
- A Slack message. Someone mentions the agent in a thread, and it picks up the work and reports back in the same thread.
- An event in your tracker or repo. An assigned issue, a label, or a PR comment.
It runs on a cloud machine, edits code, runs tests, and reports back with a pull request or a message. Triggers are now standard. What separates the products is which agents and models they run, who pays for tokens, where you steer them from, and what context they carry between runs. For more on scheduled work, see scheduled coding agents and Slack coding agents.
The autonomous and unattended agents, as of October 2026
Nori is built for the team workflow around the agent. Work starts from Slack, Discord, the web, the CLI, a cron or one-time schedule, or a webhook. Every run loads your org skillsets and shared memories, uses your own Claude or ChatGPT subscription, and runs the harness you choose: Jiro, Nori’s own engineering teammate, or Claude Code, Codex, Gemini CLI, and any other harness. Details are in where Nori fits below.
Devin (Cognition) is the reference product and the most complete single-vendor package. Automations start Devin from Slack, GitHub, GitLab, Linear, Jira, PagerDuty, cron or one-time schedules, and webhooks (Devin automations); the older Scheduled Sessions feature is now marked legacy (Devin docs). It adds memory, playbooks, skills, MCP, and hooks. SSO covers Okta, Entra ID, SAML, and OIDC, with SCIM (Devin SCIM). Outposts run session machines on infrastructure you control on Pro, Max, and Teams, though the agent “brain” always runs in Cognition’s cloud (Outposts; deployment). Plans as of October 2026: Free, Pro $20/mo, Max $200/mo, Teams from $80/mo, Enterprise via sales, with token quotas and then prepaid credits (Devin billing). It runs only Devin, does not support third-party LLM API keys, and a ChatGPT plan can pay for GPT usage but a Claude subscription cannot. See Nori vs Devin.
OpenAI Codex runs coding tasks in cloud environments you configure once; each task gets its own workspace, keeps running while your computer sleeps, and ends in a commit or pull request (Codex cloud). You start tasks from ChatGPT on web, desktop, and mobile, or from the CLI, and Codex can run scheduled tasks on a timer (Codex automations). As of October 2026, Codex is included in ChatGPT Free, Go ($8/mo), Plus ($20/mo), Pro (from $100/mo), Business ($20/user/mo billed annually), and Enterprise, with ChatGPT credits for extra usage; Plus adds cloud features such as automatic code review and the Slack integration. Using an API key instead bills at API rates and has no cloud features (Codex pricing). It is OpenAI’s agent on OpenAI’s service. (Disclosure: the Nori CLI is a fork of Codex.)
Factory sells “Droids” for the enterprise SDLC. Automations start a Droid session; webhook triggers are in Private Preview at 10 runs per hour per automation (Factory automations). Its deployment range includes isolated EU deployment, fully air-gapped mode with on-prem models, and hybrid mode through customer gateways (Factory deployment), with SSO, SAML, and SCIM on Business and Enterprise. Bring-your-own-compute (OpenCompute) is in Private Preview. Plans as of October 2026: Pro $20, Plus $100, Max $200, Teams $60/mo plus $40/seat, and custom Business and Enterprise, with rolling usage limits and per-model multipliers (Factory pricing). BYOK works in the Droid CLI and desktop app, not on the hosted web platform (Factory BYOK). Factory reportedly raised $200M at a $5B valuation in September 2026. See Nori vs Factory.
Claude Tag (Anthropic) puts Claude in your Slack channels, in public beta for Team and Enterprise plans. Routines run on a schedule, a channel watch, or a repository event, and code work lands as draft PRs through the Claude GitHub App. Channel work draws on an org usage balance with a spend limit. It is Claude-only and Slack-only (Claude Tag docs). See Nori vs Claude Tag.
GitHub Copilot cloud agent (formerly the coding agent) is the GitHub-native option. Assign an issue, or create an automation that runs repository tasks on a schedule or in response to events (GitHub docs). It is included in paid plans (Pro, Pro+, Max, Business, Enterprise, Student), not Free (Copilot plans). Since June 1, 2026, usage consumes GitHub AI Credits priced at model API rates, with each plan including credits equal to its price (GitHub blog). Starting sessions from Slack was in public preview in October 2026 (GitHub docs).
Cursor cloud agents run in isolated VMs on the Premium plan ($30/month) and up, drawing on one weekly usage pool shared with the editor and CLI; you start them from Slack and Linear (@cursor), GitHub, or the API (Cursor cloud agents). Self-hosted machines move tool execution onto your workers while the agent loop and inference stay with Cursor. Plans as of November 2026 are tiers of the X Subscription: Free and Plus $8 (no cloud agents), Premium $30, Super $100, Ultra $200, Teams from $30/user, and Enterprise (Cursor pricing). SpaceX’s acquisition of Cursor reportedly closed on August 14, 2026. See Nori vs Cursor cloud agents.
Amp Orbs are cloud VMs where Amp’s own agent runs without supervision. Webhooks give an Orb a URL, agents can schedule themselves, and you can @-mention Amp in Slack (Amp manual). Orbs bill per minute with “no Amp token fees or limits,” and you can use your ChatGPT, Claude, and other subscriptions or bring your own key (Amp pricing). See Nori vs Amp Orbs.
Tembo runs Claude Code, Codex, Cursor, Amp, and OpenCode in isolated cloud VMs, and accepts Claude, ChatGPT, and SuperGrok subscriptions. You trigger it with @Tembo in Slack, Linear, or PR comments, from new issues and PRs, the API, webhooks, or a schedule. Plans as of November 2026 are a 7-day Max trial, Pro $60/month (includes $90 of usage, up to 5 users), Max $200/month (includes $400 of usage, up to 10 users), and Enterprise, where usage covers Tembo Gateway tokens and VM time; a self-hosted or VPC deployment is offered (Tembo pricing).
Ellipsis is the config-as-code option: each agent is a YAML file in your repo that goes live on merge, triggered by cron, Slack and Linear events, or @ellipsis mentions (Ellipsis agents). As of October 2026 it bills tokens plus 10%, sandbox compute at $0.142 per vCPU-hour, and no per-seat fees, with a bring-your-own-cloud option in your AWS VPC (Ellipsis pricing).
OpenHands (All Hands) is the open-source anchor: an MIT-licensed coding agent with a hosted cloud and an enterprise product. It can also drive third-party agents such as Claude and Codex, uses your own LLM key, and can be self-hosted; OpenHands Teams (self-hosted) is $500/month for ten users plus $50/month per extra seat (OpenHands pricing). You can mention @openhands in Slack, including on self-hosted installs (OpenHands Slack). See Nori vs OpenHands.
Google Jules took a GitHub repo and branch (or an issue labeled “jules”), cloned it into a cloud VM, showed you a plan, and opened a pull request with Gemini models (Jules). Under Google’s October 29, 2026 notice, Jules has been read-only since November 11 and shuts down on December 31, 2026 (Jules changelog). See what to use after Google Jules shuts down.
Charlie Labs, which we covered in July, has shut down. It stopped new signups and remained available until October 5, 2026 (Charlie Labs).
The field keeps widening. Warp Factories runs Warp Agent, Claude Code, and Codex from Warp’s control plane. Replicas runs Claude Code, Codex, Cursor, and OpenCode in per-run Linux VMs, triggered from Linear, Slack, GitHub, GitLab, schedules, or webhooks. Open-source Broccoli (MIT) turns Linear tickets into PRs with Claude and Codex on your own Google Cloud project. The full field is in the cloud coding agents comparison.
How they compare
| Platform | Agents and models | Unattended triggers | Chat control | Where it runs | Pricing (as of October 2026) |
|---|---|---|---|---|---|
| Nori | Jiro plus any harness (Claude Code, Codex, Gemini CLI, Cursor, OpenCode, Droid, Amp, Goose, any ACP agent); your subscriptions, keys, or gateway | Cron, one-time schedules, webhooks, Slack messages | Slack, Discord, web, CLI | Nori cloud; Enterprise in your cloud or on-prem | Developer $99/mo; Team $250/mo; Enterprise custom; no token charges |
| Devin | Devin only; no third-party API keys; ChatGPT plan can pay for GPT usage | Schedules, webhooks, GitHub, Linear, Jira, PagerDuty | Slack, Teams | Cognition cloud; Dedicated VPC; Outposts for machines on your infra | Free; Pro $20; Max $200; Teams from $80/mo; Enterprise; plus usage |
| OpenAI Codex | Codex with OpenAI models | Scheduled tasks | Slack | OpenAI cloud | In ChatGPT Free to Enterprise; Plus $20; Pro from $100; credits for extra usage |
| Factory | Droid only; BYOK in CLI and desktop | Automations; webhooks in Private Preview | Slack | SaaS, EU, hybrid, air-gapped | Pro $20 to Max $200; Teams $60/mo + $40/seat; custom; plus usage limits |
| Claude Tag | Claude only | Schedule, channel watch, repo event | Slack only | Anthropic; draft PRs through the Claude GitHub App | Anthropic Team/Enterprise; org usage balance |
| GitHub Copilot cloud agent | Copilot | Schedule, repo events, assigned issues | Slack (public preview) | GitHub | Paid Copilot plans; GitHub AI Credits |
| Cursor cloud agents | Cursor agent; many models | Slack, Linear, GitHub, API | Slack | Cursor VMs; self-hosted workers for tool execution | Premium $30/mo and up; Teams $30+/user; shared usage pool |
| Amp Orbs | Amp; your subscriptions or keys | Webhook, self-scheduling | Slack | Amp cloud VMs | Per-minute Orbs; no Amp token fees |
| Tembo | Claude Code, Codex, Cursor, Amp, OpenCode | Schedule, webhook, issues, PRs, API | Slack, Linear | Cloud VMs; self-hosted or VPC offered | Pro $60/mo ($90 usage); Max $200/mo ($400 usage) |
| Ellipsis | Ellipsis agents (YAML) | Cron, Slack and Linear events | Slack, Linear | Ellipsis sandboxes; your AWS VPC | Tokens + 10%; no seat fees |
| OpenHands | OpenHands, Claude, Codex; your LLM key | Slack mentions | Slack | Open source; hosted cloud; self-host | OSS free; Teams $500/mo for 10 users, +$50/seat |
Vendor-documented capabilities and pricing as of October 2026. See sources below.
How to choose an autonomous or unattended coding agent
Six questions cut through the near-identical marketing faster than any feature grid:
- Whose agent and model do you run? Devin, Factory, Claude Tag, Amp, and Codex each run their own agent. Nori, Tembo, and OpenHands can run third-party agents. If you want to swap agents as the frontier moves, that matters more than any single benchmark.
- How do you pay for tokens? Credits, allowances, and API-rate billing put the vendor between you and the model provider. If your team already pays for Claude or ChatGPT subscriptions, check whether the product can use them.
- What can wake it up? Ticket assignment is table stakes. Schedules, webhooks, and Slack messages are what let work start with no human present. Check whether webhooks are generally available or still in preview.
- Where do you steer it? Results should land where the team already works: Slack, Discord, Linear, GitHub, or the CLI.
- Where must the machine live? Regulated teams should check for self-hosted, EU, or bring-your-own-compute options and identity integration (SAML, SCIM, OIDC).
- What does it remember? Fresh machines forget your conventions every run unless the platform loads org instructions, memories, and skills for you.
An agent is not a runtime
“Takes a task and returns a PR” is now a commodity. Before an agent ships a PR from a cloud machine, something has to stand up the machine, load the agent, connect GitHub and chat, route the model, load your org’s instructions, and give a human a way in when the agent gets stuck. Most products here pair their own agent with their own runtime and their own token billing, which ties you to one agent and one way of paying for it. Teams that want different agents for different jobs, or already pay for model subscriptions, get more from a platform that runs all of them.
Where Nori fits
Nori gives your team one workflow for autonomous and unattended coding work. It ships Jiro, Nori’s own engineering teammate, and runs any other harness you choose. Every session is a cloud machine with your repos, credentials, and org context loaded:
- Slack, Discord, web, and CLI. Mention Nori in a Slack thread and the work starts there and reports back there. Steer the same session from Discord, the web, or the CLI, and open a terminal and editor on the same machine when a human wants to step in.
- Schedules and webhooks. Cron, one-time, and webhook automations launch sessions, do the work, and report back with no human attached, with run history and approval paths.
- Memories and skillsets. Shared memories, workflow skills, and org skillsets load into every session, so agents follow your conventions from the first run.
- Your own subscriptions. We never charge for tokens. Bring your own Claude or ChatGPT subscription, API keys, or LLM gateway. No markup, no provider lock-in, and no incentive for us to make your agents burn more tokens.
- Any harness. Run Jiro, Claude Code, Codex, Gemini CLI, Cursor, OpenCode, Droid, Amp, Goose, or any ACP agent, side by side on the same team.
- Enterprise. Bring your own compute, IdP (SAML, SCIM, OIDC), LLM gateway, token provider, and memory provider, in your cloud or on-prem.
Pricing as of November 2026: a free 30-day trial (no card), Developer $99/mo, Team $250/mo, and Enterprise custom. Teams have merged more than 100,000 PRs with Nori.
Nori is the strongest fit when the work belongs to a team rather than one person: the same agents, memories, and automations serve everyone, the models run on subscriptions you already pay for, and the agent mix can change as the frontier moves without changing how the team works.
Subscribe to the newsletter
Weekly updates and long-form essays from the Nori team on AI agents and the agent ecosystem.
The bottom line
If your team wants autonomous and unattended coding work in one place, started from Slack, on a schedule, or from a webhook, with shared memories loaded, on Claude or ChatGPT subscriptions you already pay for, and with Claude Code, Codex, and Gemini CLI side by side, Nori is built for that. For a single vendor’s agent: Codex if you already pay for ChatGPT, Devin for the most complete single-vendor package, Factory for air-gapped or EU deployment, Copilot’s cloud agent for GitHub-only shops, Claude Tag for Claude-only Slack teams, Ellipsis for agent configs reviewed like code, and OpenHands for open source.
Start a free 30-day Nori trial or read the full cloud coding agents comparison.
Frequently asked questions
What is an autonomous AI software engineer?
An autonomous AI software engineer is a coding agent that takes a task, such as a ticket, an issue, or a plain-English prompt, and works it end to end on a cloud machine: it clones the repository, plans the change, writes and tests the code, and opens a pull request for review without a human driving each step. Nori, Devin, OpenAI Codex, Factory, Cursor cloud agents, and OpenHands are all versions of this idea.
What is an unattended AI coding agent?
An unattended (or background) AI coding agent is an autonomous agent that starts work without a human clicking run. A cron schedule, a webhook, or an event such as an assigned issue or a Slack message launches it; it works on a cloud machine and reports back with a pull request or a message in the thread.
Which autonomous AI software engineer is best in 2026?
It depends on how your team works. Nori is the pick for teams that want one workflow across Slack, schedules, and webhooks, with shared memories, their own Claude or ChatGPT subscriptions, and a choice of Claude Code, Codex, or Gemini CLI. OpenAI Codex is the easiest start if you already pay for ChatGPT. Devin is a complete single-vendor agent. Factory offers EU and air-gapped deployment. GitHub Copilot’s cloud agent suits GitHub-only teams. Claude Tag suits Claude-only Slack teams. OpenHands is the open-source pick.
Can AI coding agents run on a schedule, from a webhook, or from Slack?
Yes. As of October 2026, Devin automations run on cron or one-time schedules, webhooks, and Slack, GitHub, Linear, Jira, and PagerDuty events. OpenAI Codex runs scheduled tasks and has a Slack integration. Factory automations start Droid sessions, with webhook triggers in Private Preview. Claude Tag routines run on a schedule, a channel watch, or a repository event, in Slack only. GitHub Copilot automations run on a schedule or repository events. Cursor cloud agents start from Slack, Linear, GitHub, or the API. Amp Orbs take webhooks and can schedule themselves. Tembo and Ellipsis run on schedules and Slack events. Nori runs cron, one-time, and webhook automations, takes work from Slack, and is driven from Slack, Discord, the web, and the CLI.
How much do autonomous and unattended AI coding agents cost?
Most combine a plan with metered usage. As of October 2026: Devin has Free, Pro at $20/month, Max at $200/month, and Teams from $80/month, then prepaid credits. OpenAI Codex is included in ChatGPT plans from Free through Plus at $20/month and Pro from $100/month, with credits for extra usage. Factory runs from Pro at $20/month to Max at $200/month, with Teams at $60/month plus $40 per seat. Copilot plans include GitHub AI Credits equal to the plan price. Tembo has Pro at $60/month with $90 of usage and Max at $200/month with $400 of usage. Ellipsis bills tokens plus 10% with no seat fees. Cursor cloud agents need Premium ($30/month) or higher and draw on a shared weekly usage pool. Nori is Developer $99/month or Team $250/month, never charges for tokens, and has a free 30-day trial.
Can I use my own Claude or ChatGPT subscription?
Nori accepts your Claude or ChatGPT subscription, API keys, or an LLM gateway, runs any harness, and never charges for tokens. Among the others, as of October 2026: Codex runs on your ChatGPT plan. Amp and Tembo accept Claude and ChatGPT subscriptions. Devin lets a ChatGPT Go, Plus, or Pro plan pay for GPT usage but does not support Claude subscriptions or third-party API keys. Factory supports bring-your-own-key in its CLI and desktop app. Cursor meters model use against its plan’s usage pool. OpenHands uses your own LLM key.
What is the difference between an autonomous agent and an agent runtime?
An autonomous agent is the worker that reads a task and proposes code. An agent runtime is the machine, triggers, chat integrations, credentials, and org context it runs inside. Most autonomous engineering products bundle one agent with their own runtime and their own token billing. Nori ships both: Jiro, its own engineering teammate, and a runtime that runs whichever harness you choose, such as Claude Code, Codex, or Gemini CLI, with the triggers, chat control, and memories around it.
Are there open-source autonomous AI software engineers?
Yes. OpenHands by All Hands is MIT-licensed, can be self-hosted, and has a hosted cloud and enterprise product. Broccoli is an MIT-licensed project that turns Linear tickets into pull requests with Claude and Codex on your own Google Cloud project. Open-source agents give you data control and model choice, and you operate the infrastructure around them.
What happened to Charlie Labs?
Charlie Labs shut down. It stopped new signups and remained available until October 5, 2026.
Sources
- Devin self-serve billing — Cognition
- Devin ChatGPT billing — Cognition
- Devin deployment overview — Cognition
- Devin Outposts — Cognition
- Devin SCIM and SSO — Cognition
- Devin automations — Cognition
- Devin Scheduled Sessions (legacy) — Cognition
- Devin integrations — Cognition
- Codex cloud — OpenAI
- Codex pricing — OpenAI
- Codex automations — OpenAI Academy
- Factory pricing — Factory
- Factory automations — Factory
- Factory network and deployment — Factory
- Factory OpenCompute — Factory
- Factory BYOK — Factory
- Factory raises $200M at $5B valuation — HPCwire (secondary)
- Claude Tag overview — Anthropic
- About Copilot cloud agent — GitHub Docs
- Copilot plans — GitHub Docs
- Copilot usage-based billing — GitHub Blog
- Copilot cloud agent in Slack — GitHub Docs
- Cursor cloud agents — Cursor
- Cursor self-hosted machines — Cursor
- Cursor pricing — Cursor
- SpaceX completes Cursor acquisition — The Next Web (secondary)
- Orbs — Amp manual
- Amp pricing — Amp
- Tembo pricing — Tembo
- Ellipsis pricing — Ellipsis
- Ellipsis agents — Ellipsis
- OpenHands pricing — All Hands
- OpenHands Slack integration — All Hands
- Jules — Google
- Jules changelog — Google
- Jules usage limits — Google
- Google AI plans — Google
- Charlie is shutting down — Charlie Labs
- Open infrastructure for building a software factory — Warp
- Replicas — background coding agents
- Broccoli — open-source ticket-to-PR agent