Gemini 3.6 Flash vs GPT-5.2
Google against OpenAI, costed on the same three workloads. Gemini 3.6 Flash is cheaper on all three workloads.
Monthly bill, side by side
| Workload | Gemini 3.6 Flash | GPT-5.2 | Difference |
|---|---|---|---|
| Support chatbot 2,500 six-turn conversations a day |
$3,021 | $9,174 | $6,154 (3×) |
| Document summariser 1,000 single-shot summaries a day |
$364 | $977 | $612 (2.7×) |
| Coding agent 200 twelve-turn sessions a day |
$2,106 | $5,859 | $3,753 (2.8×) |
No prompt caching and no batch discount on either side, so the comparison is like for like. Both figures assume 30 days a month.
The rate cards
| Gemini 3.6 Flash | GPT-5.2 | |
|---|---|---|
| Input, per 1M tokens | $0.75 (Promotional rate until 2026-12-31 — standard rate $1.5/M) | $1.75 |
| Output, per 1M tokens | $3.75 (Promotional rate until 2026-12-31 — standard rate $7.5/M) | $14 |
| Cached input | $0.075 (Promotional rate until 2026-12-31 — standard rate $0.15/M) | $0.175 |
| Context window | 1.05M | 272K |
| Maximum output | 66K | 128K |
| Reasoning model | yes | yes |
| Vision | yes | yes |
Run this comparison on your own numbers
Full detail: Gemini 3.6 Flash pricing · GPT-5.2 pricing