openinterpreterTerminal coding agent optimized for low-cost open models — a Rust Codex fork with switchable harness emulation (claude-code, kimi-code, qwen-code…), ACP support and Codex SDK compatibility.
Why switchTwo Rust terminal agents attacking the same bill from opposite ends: jcode cuts footprint (~28MB per session against 140-390MB), openinterpreter cuts model cost by targeting cheap open models with switchable providers.
Full comparison → tauHugging Face's Python port of Pi's minimalist coding agent: a real terminal agent with TUI, sessions and skills — built to be READ, with a clean brain/environment/frontend separation.
Why switchBoth are lean open coding agents rejecting harness bloat: tau is a minimalist Python port of Pi with TUI and skills; jcode is a Rust binary optimized to the byte — pick Python hackability or raw footprint.
Full comparison → openhumanRust agent harness built for fleets: desktop app, terminal or embeddable library, with built-in memory, browser/desktop control, 119 OAuth apps and 14 channels — 500 agents on a $10 VPS.
Why switchBoth are Rust harnesses that sell footprint — tens of MB per session so many run in parallel; jcode stays a coding agent, OpenHuman adds memory, channels, browser control and an embeddable library.
Full comparison → mini-swe-agentThe 100-line agent from the SWE-bench team: >74% on SWE-bench Verified with no tools but bash, no config sprawl — the reference minimal harness, adopted by Meta, NVIDIA and Ramp.
Why switchTwo minimalisms: mini-swe-agent is pedagogical — 100 lines, bash only, read it in one sitting; jcode is engineered — full harness features with the resource profile of a grep.
Full comparison →