Head to head
Google: Gemini 3.8 Flash vs OpenAI: GPT-5.6 Luna
Live prices, context windows, capabilities, and measured quality — updated continuously.
Side by side
| Metric | Google: Gemini 3.8 Flash Google Ai Studio | OpenAI: GPT-5.6 Luna Openai |
|---|---|---|
| Input $/M tokens | $0.750 | $0.200 |
| Output $/M tokens | $3.75 | $1.20 |
| Context window | 1.0M | 1.1M |
| Max output | 66K | 128K |
| WM Score | 88 | 85 |
| Benchmark quality | 86.9 | 80.2 |
| Hosts serving it | 2 | 3 |
| Throughput (tok/s) | 149 | 96 |
| Latency (ms) | 2460 | 1036 |
| Vision | ||
| Tools | ||
| Reasoning | ||
| JSON mode |
Which should you pick?
OpenAI: GPT-5.6 Luna costs less per million input tokens — about 3.8× cheaper than Google: Gemini 3.8 Flash. Google: Gemini 3.8 Flash scores higher on measured quality. OpenAI: GPT-5.6 Luna takes a larger context window, which matters for long documents or big codebases. Prices here are live list prices; the same model can cost less on a different host, so check the model page for the cheapest offering.