Verdict
Built from the measurements below
- Accuracy is too close to call: Claude Fable 5.1 scored 81.2% and Gemini 3.1 Pro Preview 77.5% on the same 80 questions, a gap within what chance alone could produce.
- They cost about the same per question ($0.016 vs $0.017).
- Gemini 3.1 Pro Preview generates text 1.8× faster: 124 tokens per second against 70.
- Claude Fable 5.1 starts answering sooner: 2.1 s to the first token, against 3.4 s for Gemini 3.1 Pro Preview.
Which should you choose?
Choose Claude Fable 5.1 for the lower cost per correct answer ($0.02 against $0.022), a quicker first token, shorter outputs (about 238 tokens per question against 1,382) and more hosts to choose from on OpenRouter (4 against 2). Choose Gemini 3.1 Pro Preview for faster streaming. The widest subject gap is in law: Claude Fable 5.1 answered 5 of 6 correctly, Gemini 3.1 Pro Preview 3 of 6. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.
Accuracy gap +3.7 points (z = 0.59). We call a winner only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.