o3 vs GPT-5.2: cost compared
Last verified
GPT-5.2 is 13% cheaper on input and -75% cheaper on output than o3. Whether that makes it the right choice depends entirely on whether it does your task well enough — this page gives you the cost side of that trade, precisely.
Rates side by side
| $ per million tokens | o3 | GPT-5.2 |
|---|---|---|
| Input | $2.00 | $1.75 |
| Output | $8.00 | $14.00 |
| Cached input | $0.50 | $0.17 |
| Output : input ratio | 4.0× | 8.0× |
What the difference is worth
Monthly cost for the same workload on each, without caching:
| Workload | o3 | GPT-5.2 | Difference |
|---|---|---|---|
| Support chatbot 1,000 conversations/month · 2,000 in + 500 out each | $8.00 | $10.50 | $-2.5000 saved |
| RAG search 10,000 queries/month · 8,000 in + 400 out each | $192 | $196 | $-4.0000 saved |
| Bulk extraction 10,000 documents · 3,000 in + 300 out each | $84.00 | $94.50 | $-10.5000 saved |
| Coding agent 100 runs/month · 400,000 in + 20,000 out each | $96.00 | $98.00 | $-2.0000 saved |
Which to use
GPT-5.2 if the task is well-specified and mechanical — extraction, classification, formatting, routine transformation. The saving is real and compounds with volume.
o3 if the task involves genuine reasoning, long-horizon planning, or work where a wrong answer is expensive to catch. Paying 1.1× more on input is trivial compared to the cost of shipping a bad result.
Both are OpenAI models, so switching is usually a one-line change and the tokenizer is the same — which makes this a genuinely easy experiment to run. Test on your own workload before deciding.
Don't skip caching
Cached input costs $0.50 on o3 and $0.17 on GPT-5.2. If your prompts share a stable prefix, caching on the more expensive model can beat switching to the cheaper one outright — worth checking before you migrate anything.
Sources
OpenAI rates verified 2026-08-05 (source). OpenAI rates verified 2026-08-05 (source).