A full OpenCode agent team on free models

Run every OpenCode agent on a free model with fallbacks. oh-my-opencode-slim users on zero budget.

Share on 𝕏 ↗ Compare with… Read the source ↗

what to steal

Give every agent a fallback array so a rate limit never stops the run.

Our takeaway from the source, not a quote.

Drawn from this card's sourced fields. Amber = older model named by the source. Rose = cut or replaced.

michyamrane-omo-slim-free.toml

# Run every OpenCode agent on a free model with fallbacks
# source: https://michyamrane.github.io/blog/implementing-oh-my-opencode-slim-in-opencode/

[rig]
who     = "oh-my-opencode-slim users on zero budget"
author  = "mIcHyAmRaNe"
date    = "not stated"

harness = ["OpenCode", "OMO (Oh My OpenAgent)"]
models  = ["Free models (per-agent fallback arrays)"]

[keep]
  - Agents: orchestrator, oracle, explorer, librarian, designer, fixer
  - Fallback arrays per agent

Source: https://michyamrane.github.io/blog/implementing-oh-my-opencode-slim-in-opencode/ · checked 2026-10-14

OpenCode + OMO, 100% local

steal → Pair a local agent with a self-hosted forge like Gitea and nothing leaves the box.

OpenCodeOMO (Oh My OpenAgent) opencode/minimax-m3-fr

by Veladan

~/rigs/rhcwlq89-omo-routing.toml2026-02-08

Each OMO agent gets its own model

steal → Route by role: put the expensive model only where reasoning matters.

OpenCodeOMO (Oh My OpenAgent) Claude Opus 4.5 HighGPT 5.2 Codex

by rhcwlq89

~/rigs/joekarlsson-opencode-litellm.toml2026-10

Broke up with Claude Code: open-weight models at about two cents a task

steal → Put every model call behind one gateway with per-app budgets, and force a test run before 'done'.

OpenCodeClaude Code DeepSeek V4.1 FlashClaude Claude Code as default

by Joe Karlsson