GPT-5.6 Luna vs Qwen3 Coder 480B A35B
| GPT-5.6 Luna | Qwen3 Coder 480B A35B | |
|---|---|---|
| Coding index | 61.5 | 50.0 |
| Benchmarks behind it | Partial, 2 benchmarks | Partial, 2 benchmarks |
| Output $/Mtok | $1.20 | $1.00 |
| Cost per solved task | $0.04 | Not computable |
| Context window | 1.05M | 262K |