Hi, @Mohamed squad mode reuses shared context + compute, so the same assistant isn’t spun up from scratch each time—less duplicated tokens, fewer tool calls, lower cost. If you want, I can help you redesign your setup to get squad pricing with solo behavior.