Verdict
Built from the measurements below
- Accuracy is too close to call: DeepSeek V4 Flash Vision Exp scored 71.2% and GPT-6 Luna 72.5% on the same 80 questions, a gap within what chance alone could produce.
- GPT-6 Luna is 4.3× cheaper per question as billed ($0.00025 vs $0.0011), reasoning tokens included.
- DeepSeek V4 Flash Vision Exp generates text 1.2× faster: 170 tokens per second against 137.
- DeepSeek V4 Flash Vision Exp starts answering sooner: 0.7 s to the first token, against 1.2 s for GPT-6 Luna.
Which should you choose?
Choose DeepSeek V4 Flash Vision Exp for faster streaming, a quicker first token and open weights you can run yourself. Choose GPT-6 Luna for the lower cost per correct answer ($0.00034 against $0.0015) and shorter outputs (about 431 tokens per question against 1,301). The widest subject gap is in medicine: GPT-6 Luna answered 6 of 11 correctly, DeepSeek V4 Flash Vision Exp 4 of 11. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.
Accuracy gap -1.2 points (z = -0.18). We call a winner only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.