mini-SWE-agentvsZero
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
mini-SWE-agent
SWE-agent (Princeton and Stanford) · Terminal agent
#35OSS
Panel
6.92 spec wins
- Reliability
- 6.3
- Usefulness
- 5.5
- Cost
- 8.8
- Longevity
- 7.0
“One hundred lines of Python, which is fewer than most competitors spend on their pricing page.”
Zero
Gitlawb · Terminal agent
#6OSSMCP
Panel
7.04 spec wins
- Reliability
- 6.7
- Usefulness
- 7.0
- Cost
- 8.2
- Longevity
- 6.2
“It accepts image input in a terminal, a sentence that would have ended a career in 2019.”
Spec by spec
| Spec | mini-SWE-agent | Zero |
|---|---|---|
| Architecture | ||
| Category | Terminal agent | Terminal agent |
| Runs | local, sandbox | local, sandbox |
| Platforms | macos, linux | macos, linux, windows |
| Context window | not documented | not documented |
| Protocols | ||
| MCP client | No | Yes |
| MCP server | No | Yes |
| Capabilities | ||
| Runs terminal commands | Yes | Yes |
| Multi-file edits | Yes | Yes |
| Git operations | Yes | Yes |
| Browser control | No | YesBrowser and terminal control helpers ship with the npm and release builds; source builds need those helper binaries on PATH or configured in local-control settings. |
| Sandboxed execution | Yes | NoZero ships a native Linux sandbox helper using seccomp rather than Docker; source builds must build the `zero-linux-sandbox` helper separately. |
| Multi-agent orchestration | No | Yes |
| Headless / CI mode | Yes | Yes`zero exec` returns exit codes and has a documented GitHub Action recipe. |
| Models | ||
| Backbone | any via litellm | OpenAI, Anthropic, Gemini, Groq, OpenRouter, DeepSeek, Mistral, xAI, Qwen, Kimi, GitHub Models, Fireworks, MiniMax, Ollama, LM Studio |
| Bring your own model | Yes | Yes |
| Local models | Yes | Yes |
| Cost | ||
| Pricing model | byok | byok |
| Starts at | n/a | $0/mo |
| Free tier | Yes | Yes |
| Bring your own key | Yes | Yes |
| Openness | ||
| Open source | Yes | Yes |
| License | MIT | MIT |
| GitHub stars | 8,145 | 1,688 |
Which one would each critic pick
| Critic | mini-SWE-agent | Zero | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 6.8 | 7.3 | Zero — Pick it if you want a terminal agent that lets you change your mind mid-task; pick Crush if you want the same shape with a longer track record behind it. |
| El Crítico | 6.5 | 6.8 | Zero — Package builds ship the browser and terminal control helpers; a source build needs those binaries on PATH and expects you to build the Linux sandbox helper separately. |
| El Profesor | 8.0 | 7.3 | mini-SWE-agent — The reference harness for the SWE-bench bash-only leaderboard: one bash action per turn via subprocess.run, a linear history, no tool-calling API, and Gemini 3 Pro reported above 74% on Verified. |
| La Inversora | 6.8 | 5.5 | mini-SWE-agent — No company, a Princeton and Stanford lab with 6,938 stars; the funding is grants and the exit is a paper, which is more stable than half the cap tables on this board. |
| La Jefa | 5.0 | 6.8 | Zero — Nothing to license for sixty, and it is the rare tool that returns meaningful exit codes with a documented continuous integration recipe, so it can gate a build. |
| El Hacker | 8.5 | 8.5 | no preference |
Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.