I've been using opencode for a few months and really like it, both from a UX and a results perspective.
It started getting increasingly flaky with Anthropic's API recently, so I switched back to Claude Code for a couple of days. Oh my, what a night and day difference. Tokens, MCP use, everything.
For anyone reading at OpenAI, your support for OpenCode is the reason I now pay you 200 bucks a month instead.
I've been paying OpenAI 200 bucks a month for what feels like forever by now, but used OpenCode for the first time yesterday, been using Codex (and Claude Code from time to time, to see if they've caught up with Codex) since then.
But I don't use MCP, don't need anything complicated, and not sure what OpenCode actually offers on top. The UI is slightly nicer (but oh so much heavier resource usage), both projects source code seems vibecoded and the architecture is held together with hopes and dreams, but in reality, minor difference really.
Also, didn't find a way in OpenCode to do the "Fast Mode" that Codex has available, is that just not possible or am I missing some setting? Not Codex-Spark but the mode that toggles faster inference.
It started getting increasingly flaky with Anthropic's API recently, so I switched back to Claude Code for a couple of days. Oh my, what a night and day difference. Tokens, MCP use, everything.
For anyone reading at OpenAI, your support for OpenCode is the reason I now pay you 200 bucks a month instead.