Skip to content
ADE Index

Codex CLI vs Pi vs SWE-agent

Up to four tools, row by row. We do not guess: a cell we have not checked says so.

This page compares the three CLI agents only. Codex has 5 other listings, each with its own page.

How they differ

SWE-agent has been superseded by mini-SWE-agent: Most of our current development effort is on mini-swe-agent, > which has superseded SWE-agent. github.com

Price
Pick this if…

Paraphrased from each maker’s own description. Select a line for the quote.

Checked for Codex CLI and Pi, not yet for SWE-agent: Systems (Codex CLI: , , , ; Pi: , , , )

Not yet checked for any of these tools: agents it runs.

Dotted values open their quote.

What each can do

Select an answer to see the maker’s sentence, or what we searched.

Only 1 capability is checked for SWE-agent so far, so there is little to compare yet. 21 capabilities are checked for at least two of these tools. 6 get different answers, among them Approval modes, Subagents, Plan mode and Notifications; 2 differ only in a limit. 10 others are checked for one tool only.

Differences only8

Showing 8 that differ. Hidden: 13 where the checked answers match and 10 checked for one tool only.

  • Yesthe maker documents it
  • With add-onAdd-onsomething the maker publishes that you install or run
  • No, per makerNothe maker says it does not
  • *its limit or add-on is named under the row
  • Emptynot yet checked

Different answers6

Approval modesNot yet checked: SWE-agent
Not yet checked for SWE-agent
SubagentsNot yet checked: SWE-agent
Not built in; Subagent example
Not yet checked for SWE-agent
Plan modeNot yet checked: SWE-agent
Not built in; Plan Mode example
Not yet checked for SWE-agent
Read-only modeNot yet checked: SWE-agent
Not built in; Plan Mode example
Not yet checked for SWE-agent
NotificationsNot yet checked: SWE-agent
Maker's notify example
Not yet checked for SWE-agent
OS-level sandboxNot yet checked: SWE-agent
Linux/WSL2 needs bubblewrap
Gondolin micro-VM example
Not yet checked for SWE-agent

Differ only in a limit2

Runs in a containerNot yet checked: Codex CLI
Not yet checked for Codex CLI
The repo's Docker recipe
Needs Docker
Image inputNot yet checked: SWE-agent
Needs a compatible terminal
Not yet checked for SWE-agent
Not yet checked: , ,

6 capabilities are not yet checked for any of these tools and are not shown. Each answer is about the CLI agent in its column. Something the maker publishes separately, such as an editor plugin or a GitHub Action, shows as “With add-on”, even when it has its own listing here. How capabilities are checked

About these tools

Codex CLI

Pi

SWE-agent

How to read the Index score

At a glance

Index score out of 100

Codex CLI91Pi78SWE-agent8

Mindshare HN mentions, 90 days

Codex CLI4,237Pi1,310SWE-agent1

Adoption GitHub stars, not scored

Codex CLI136kPi123kSWE-agent20.8k

Release days last 90 days

Codex CLI32Pi19SWE-agent0

Lowest monthly plan billed monthly, lower is cheaper

Codex CLI$20/moPiNot yet checkedSWE-agentNot yet checked

Platforms systems and surfaces

Codex CLI4Pi4SWE-agent1

Readers who say they use it

Already use any of these? Tick it. Several is fine.

Votes can only raise a score: up to 20% of it reflects votes, graded on a curve against the most-voted tool. The square-root curve has a floor of 100 effective votes. Votes decay with a 90-day half-life. Maker self-votes do not count. Counts below 10 are hidden for display only. A vote is a signed-in reader ticking “I use this”. Since votes joined the score on 18 Nov 2026, the score is 80% mindshare plus 20% of the greater of mindshare and the vote score. GitHub stars do not affect the Index score. Mindshare uses a log scale; votes use the square-root curve and a 100-vote floor. Ranks use unrounded scores. Mindshare and adoption bars are log-scaled; the others are linear within this comparison. How the score is built

Differences only
3 of 4
Codex CLIOpenAI · CLI agent
PiEarendil Works · CLI agent
SWE-agentDeprecatedPrinceton University and Stanford University · CLI agent
Interface
Codex CLI explaining a codebase, with an updated plan checklist
pi startup listing loaded context, skills and extensions, then reading docs and running bash
What it is
Coding agent from OpenAI that runs locally on your computer
Minimal, extensible agent harness that you can make your own
SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice.
Category
CLI agentCodex family
CLI agent
CLI agentDeprecated
Pricing
Starts at (billing shown)
Lowest monthly plan
Not yet checked
Not yet checked
Free tier
Not yet checked
Plans
Runs on
Linux
Yes
Yes
Not yet checked
Windows
Yes
Yes
Not yet checked
macOS
Yes
Yes
Not yet checked
Models and agents
Models
GPT-6.2 AstraGPT-6.2 SolGPT-6.1 TerraGPT-6.1 SolGPT-6 Luna
Not yet checked
GPT-4oClaude Sonnet 4
Source and activity
Source
GitHub stars
Most
Release days, last 90 days
32Most
19
0
Release days per week
None in 26 weeks
Latest release, any channel
v1.5.2 10 Nov 2026
v1.1.0 22 May 2025
Index score
Score and rank
91 #2 Score 90.7; ranked on the unrounded score
78 #4 Score 77.9; ranked on the unrounded score
8 #153 DeprecatedScore 7.5; ranked on the unrounded score
Mindshare
914,237 Hacker News mentions in the last 90 days
781,310 Hacker News mentions in the last 90 days
81 Hacker News mention in the last 90 days
Adoption, not scored
97135,675 stars, read 18 Nov 2026
96122,575 stars, read 18 Nov 2026
8120,825 stars, read 18 Nov 2026
Rank history
First ranked 18 Nov 2026
First ranked 18 Nov 2026
First ranked 18 Nov 2026
Full page

Watch Codex CLI vs Pi vs SWE-agent on YouTube

More comparisons

Build your own