OpenAI cut list prices for GPT-5.6 Sol: input down 20%, output down 33%. Vercel's existing 50% AI Gateway discount now stacks on top of those lower numbers through September 18. On the Default tier, that puts effective pricing at $2.00 per million input tokens and $10.00 per million output tokens, down from $2.50 and $15.00. Flex and Priority tiers move by the same proportion.
No code changes required. The model ID, openai/gpt-5.6-sol, is unchanged, so existing requests automatically bill at the new rate. The discount covers all service tiers, cached tokens, cache writes, and long-context requests above 272K tokens. BYOK users are exempt and bill directly at their own OpenAI rate.
The piece is worth reading in full for the coding agent integration detail: run vercel ai-gateway coding-agents setup, select openai/gpt-5.6-sol in model settings, and your agents route through AI Gateway at these rates. The full pricing table, including US regional rates and cache write costs, lives on Vercel's pricing page.
[READ ORIGINAL →]