Gemini 4 Argon burns $10 on one coding task in user test
A user says Argon used ~4M tokens and $10 of credits on an 11-minute coding task without finishing, while Claude Opus 5.5 Medium passed all 90 tests with ~240K tokens.
GEMINI 4 ARGON JUST BURNED $10 ON ONE CODING TASK 💀 And somehow people still call it “cheaper” than other frontier models. I thought they were exaggerating… So I tested it myself. I gave Argon a simple coding task that took around 11 minutes. It used ~4 MILLION tokens 😭 Same task with Claude Opus 5.5 Medium: ~240K read tokens Opus finished it and passed all 90 final tests. Then I thought maybe Antigravity was the problem. So I added $10 of Gemini CLI credits and ran it again. Argon burned through the entire $10… Still couldn’t finish. 💀 So now I genuinely want to know: What are Argon users doing differently? 👀 Different harness? Better context management? A trick to stop repeated file reads? If you’ve figured out how to make Argon actually cost-efficient, please share your workflow.