Opus 5 used $600+ of credits and made this entire 10 minute playabale ghost of tsushima browser game after running a gauntlet loop for 5+ hours
Like the previous tests I've posted (full video soon), this did NOT use the Red Sands OS repo but instead started completely from
Muse Spark 1.3 is now in Codex!!
OpenCode's free tier runs it for nothing, and $10 on OpenCode Go gets you over 45,300 requests. It's basically Opus 5 tier too. There's a catch though. Meta may use your inputs and outputs for training, which is the whole reason the tier is
Coding is where it gets more interesting.
On Terminal Bench, which tests models on real terminal tasks: Quasar: 69.3 It beat Mistral Medium 3.5 by 18.7 points and Nemotron 3 Ultra by 15.4. Claude Opus 5 still leads at 89.1. Want the SOP? DM me.
Gemini 3.8 Flash vs Opus 5
You have reached the end of the archive
All of Claude Opus 5