Token accounting corrected for mixture-of-experts models

We were counting tokens against the total parameter count rather than the active experts, which overcharged Mixtral and DeepSeek V3 requests by roughly eight per cent. Affected accounts have been credited automatically; nobody needs to file anything.