OpenAI cut the o3 reasoning model price by 80%. Same model, 5× cheaper.
What this means for you
Input tokens went from $10 to $2 per million. Output from $40 to $8 per million. OpenAI optimized their inference infrastructure — the model itself is identical. Effective immediately for all API users.
Bottom line impact
If you spend $1,000/month on o3, you now spend $200. For every $10K of o3 spend, you save $8,000/year automatically. No action required.
Should you care?
✓ Yes
Yes — if you use o3 at any volume. Savings apply immediately. Also worth reconsidering o3 for use cases you previously ruled out as too expensive.
Nothing to do. New pricing is applied automatically. Update cost forecasts and reconsider o3 for workloads previously ruled out on cost.
$800/mo per $1K of o3 spend
Model quality and latency unchanged — same model, optimized infrastructure.
No policy changes.
OpenAI API Pricing
https://openai.com/api/pricing/
VentureBeat coverage
https://venturebeat.com/ai/openai-announces-80-price-drop-for-o3-its-most-powerful-reasoning-model
Mardii links only to official provider pages.
Free account. Follow the providers you care about. Unsubscribe anytime.