Verdict
Built from the measurements below
- Accuracy is too close to call: Gemini 3.5 Flash Lite scored 75.0% and Gemini 4 Argon 84.8% on the same 80 questions, a gap within what chance alone could produce.
- Gemini 3.5 Flash Lite is 1.9× cheaper per question as billed ($0.0034 vs $0.0063), reasoning tokens included.
- Gemini 3.5 Flash Lite generates text 5.9× faster: 409 tokens per second against 69.
- Gemini 3.5 Flash Lite starts answering sooner: 1.1 s to the first token, against 3.3 s for Gemini 4 Argon.
Which should you choose?
Choose Gemini 3.5 Flash Lite for the lower cost per correct answer ($0.0045 against $0.0074), faster streaming and a quicker first token. Choose Gemini 4 Argon for shorter outputs (about 558 tokens per question against 1,311). The widest subject gap is in law: Gemini 4 Argon answered 6 of 6 correctly, Gemini 3.5 Flash Lite 4 of 6. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.
Accuracy gap -9.8 points (z = -1.54). We call a winner only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.