Verdict
Built from the measurements below
- Accuracy is too close to call: gpt-oss-120b scored 58.8% and MiniMax M3 57.5% on the same 80 questions, a gap within what chance alone could produce.
- gpt-oss-120b is 20× cheaper per question as billed ($0.00013 vs $0.0026), reasoning tokens included.
- MiniMax M3 generates text 1.5× faster: 121 tokens per second against 82.
- MiniMax M3 accepts more context: 1M tokens against 131k.
Which should you choose?
Choose gpt-oss-120b for the lower cost per correct answer ($0.00021 against $0.0045), shorter outputs (about 546 tokens per question against 2,129) and more hosts to choose from on OpenRouter (23 against 12). Choose MiniMax M3 for faster streaming and a 1M-token context window. The widest subject gap is in law: MiniMax M3 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 +1.3 points (z = 0.16). We call a winner only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.