Repository navigation
0.3.3
Fixed
- Generation speed counts the time a model spends writing a tool call's
arguments. Those tokens were counted but their streaming time was not, so
a step that wrote a file showed several times its real speed (3,672 tokens
"at 162.9 tok/s" against the engine's 36.4), and the Session average was
inflated with it. Affects OpenCode's figures, not an engine's own. On
MTPLX, a turn that wrote a file now gives 36.4 tok/s from OpenCode's
figures against the engine's 36.1.
Install: opencode plugin add @banburist/opencode-headsup