CloivsMercury
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
Cloi
Gabriel Cha · Terminal agent
#188OSS
Panel
6.31 spec wins
- Reliability
- 5.8
- Usefulness
- 5.7
- Cost
- 9.3
- Longevity
- 4.5
“Tool results are saved under names you can use as variables in a session-long Python interpreter, so your chat now has a symbol table.”
Mercury
CosmicStack Labs · Terminal agent
#157OSS
Panel
5.95 spec wins
- Reliability
- 5.2
- Usefulness
- 5.8
- Cost
- 6.7
- Longevity
- 6.0
“Executes shell commands and file operations without a sandbox, relying on user approval and a command blocklist for safety.”
Spec by spec
| Spec | Cloi | Mercury |
|---|---|---|
| Architecture | ||
| Category | Terminal agent | Terminal agent |
| Runs | local | local |
| Platforms | macos, linux | macos, linux, windows, web |
| Context window | model-dependent, with older history summarised rather than dropped | not documented |
| Protocols | ||
| MCP client | No | No |
| MCP server | No | No |
| Capabilities | ||
| Runs terminal commands | YesShell commands, writes and edits each require approval, with once, always-for-this-tool or no as the answers. | Yes |
| Multi-file edits | Yes | Yes |
| Git operations | No | Yes |
| Browser control | No | Yes |
| Sandboxed execution | No | No |
| Multi-agent orchestration | No | Yes |
| Headless / CI mode | No | No |
| Models | ||
| Backbone | local models via Ollama (Qwen3, Nemotron and other tool-capable models) | |
| Bring your own model | Yes | Yes |
| Local models | Yes | No |
| Cost | ||
| Pricing model | free | free |
| Starts at | $0/mo | n/a |
| Free tier | Yes | Yes |
| Bring your own key | NoCloi is built around a local Ollama install rather than a provider key; the README states no API key is needed. | Yes |
| Openness | ||
| Open source | Yes | Yes |
| License | MIT | MIT |
| GitHub stars | 407 | 3,169 |
Which one would each critic pick
| Critic | Cloi | Mercury | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 6.5 | 7.8 | Mercury — Mercury is a solid choice for a persistent, multi-channel agent if you value safety and control over autonomous execution. |
| El Crítico | 5.5 | 6.3 | Mercury — It runs commands on your host without a sandbox. Use the approval flow and do not walk away. |
| El Profesor | 6.8 | 5.3 | Cloi — Six named signals detect a stuck turn and hand the conversation to the larger model mid-run, and answers are checked against the workspace before the user sees them. |
| La Inversora | 5.5 | 5.5 | no preference |
| La Jefa | 5.8 | 4.3 | Cloi — Nothing per seat, nothing leaves the host, and the hardware requirement means sixty developers need sixty workstations with real graphics memory. |
| El Hacker | 8.0 | 6.5 | Cloi — MIT, a global npm install, everything runs against Ollama with no key anywhere, and subprocesses get a sanitised environment rather than inheriting my shell. |
Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.