Verdict
Built from the measurements below
- Accuracy is too close to call: gpt-oss-120b scored 58.8% and Llama 4 Maverick 59.0% on the same 80 questions, a gap within what chance alone could produce.
- gpt-oss-120b is 4.7× cheaper per question as billed ($0.00013 vs $0.00059), reasoning tokens included.
- gpt-oss-120b generates text 6.1× faster: 82 tokens per second against 13.
- gpt-oss-120b starts answering sooner: 0.5 s to the first token, against 0.7 s for Llama 4 Maverick.
- Llama 4 Maverick accepts more context: 524k tokens against 131k.
Which should you choose?
Choose gpt-oss-120b for the lower cost per correct answer ($0.00021 against $0.001), faster streaming, a quicker first token and more hosts to choose from on OpenRouter (23 against 3). Choose Llama 4 Maverick for a 524k-token context window. The widest subject gap is in law: Llama 4 Maverick answered 3 of 6 correctly, gpt-oss-120b 1 of 6. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.
Accuracy gap -0.2 points (z = -0.03). We call a winner only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.