Omnisend
October 8, 2026Base fee$16$11-30%
Price Tracker
AI model & API price changes
A live changelog of every token price cut and increase across AI models, plus pricing updates for developer APIs and cloud tools. Prices are checked automatically against official vendor pricing pages.
Changes tracked
188
Price cuts
100
Price increases
81
Last change
October 8, 2026
Biggest recent cut: o3 input $10 → $2.00 (-80%)
October 8, 20268 modelsAI model price changes · Oct 8
Cached input /1M$0.14$0.07-54%Output /1M$7.00$3.39-52%Cached input /1M$0.26$0.14-46%Cached input /1M$0.07$0.05-26%
Input /1M$0.95$0.47-51%Cached input /1M$0.16$0.10-39%Output /1M$4.00$2.45-39%Input /1M$0.47$0.44-6%
Input /1M$0.28$0.26-8%
Cached input /1M$0.04$0.03-26%Cached input /1M$0.03$0.02-18%Input /1M$0.06$0.05-18%Input /1M$0.07$0.06-15%Output /1M$0.20$0.17-15%Output /1M$0.16$0.14-12%Output /1M$0.17$0.16-6%
Input /1M$0.22$0.10-55%Output /1M$0.50$0.32-36%
Cached input /1M$0.80$0.33-59%Output /1M$15$12-18%Output /1M$15$13-13%
Output /1M$0.45$0.40-11%
October 7, 20266 modelsAI model price changes · Oct 7
Input /1M$0.41$0.17-58%Cached input /1M$0.26$0.16-38%Input /1M$0.65$0.41-37%
Cached input /1M$0.06$0.02-67%Input /1M$0.10$0.05-49%
Output /1M$4.40$3.04-31%Input /1M$1.40$0.97-31%Cached input /1M$0.26$0.18-31%
Input /1M$0.55$0.45-18%Output /1M$3.50$3.00-14%Cached input /1M$0.23$0.22-2%
Output /1M$1.50$1.20-20%Input /1M$0.35$0.30-14%
Output /1M$3.25$2.00-38%Input /1M$0.32$0.30-6%
October 5, 20265 modelsAI model price changes · Oct 5
Input /1M$4.00$2.00-50%Output /1M$20$10-50%Cached input /1M$0.40$0.20-50%
Input /1M$0.13$0.08-38%Output /1M$0.53$0.33-38%Cached input /1M$0.03$0.02-38%
Input /1M$0.03$0.02-27%Input /1M$0.02$0.02-19%Cached input /1M$0.02$0.02-19%Input /1M$0.02$0.02-15%
Output /1M$3.00$2.55-15%
Cached input /1M$0.04$0.04-10%Input /1M$0.83$0.75-10%Output /1M$2.50$2.25-10%
October 3, 202610 modelsAI model price changes · Oct 3
Output /1M$1.40$1.10-21%Input /1M$0.32$0.26-20%Output /1M$1.10$0.89-19%
Output /1M$1.04$0.79-24%Cached input /1M$0.01$0.0088-24%Input /1M$0.35$0.26-24%Input /1M$0.26$0.22-17%
Input /1M$0.30$0.21-30%Output /1M$1.20$0.84-30%Cached input /1M$0.06$0.04-30%
Cached input /1M$0.0070$0.0069-2%Output /1M$0.66$0.65-2%Input /1M$0.22$0.22-2%
Input /1M$0.60$0.50-17%Cached input /1M$0.12$0.10-17%Output /1M$2.40$2.20-8%
Cached input /1M$0.04$0.02-59%Input /1M$0.51$0.21-59%Output /1M$1.02$0.42-59%Input /1M$0.60$0.51-15%Output /1M$1.20$1.02-15%Cached input /1M$0.05$0.04-15%
Input /1M$0.05$0.03-40%Output /1M$0.09$0.06-40%Cached input /1M$0.0094$0.0056-40%
Cached input /1M$0.06$0.05-9%
Input /1M$0.31$0.15-52%Output /1M$1.25$1.00-20%
Input /1M$0.10$0.05-52%Output /1M$0.30$0.19-36%
September 25, 20265 modelsAI model price changes · Sep 25
Input /1M$4.00$2.00-50%Output /1M$20$10-50%Cached input /1M$0.40$0.20-50%
o3
September 25, 2026Input /1M$10$2.00-80%Cached input /1M$1.00$0.50-50%
Cached input /1M$1.00$0.50-50%
Cached input /1M$0.31$0.13-60%Cached input /1M$0.16$0.13-20%
Cached input /1M$0.07$0.03-60%Cached input /1M$0.04$0.03-20%Input /1M$0.35$0.30-14%
September 18, 20261 modelAI model price changes · Sep 18
Neon
September 18, 2026Base plan$29$19-34%
August 27, 20261 modelAI model price changes · Aug 27
Resend
August 27, 2026Overage rate$0.0010$0.0009-10%
August 19, 20261 modelAI model price changes · Aug 19
llama-3-3-70b
August 19, 2026Input /1M$0.88$0.59-33%
August 11, 20261 modelAI model price changes · Aug 11
Pinecone
August 11, 2026Base plan$70$50-29%
July 30, 20261 modelAI model price changes · Jul 30
Output /1M$0.75$0.60-20%
July 21, 20261 modelAI model price changes · Jul 21
Clerk
July 21, 2026Overage rate$0.03$0.02-20%
July 12, 20261 modelAI model price changes · Jul 12
Input /1M$3.00$2.00-33%
How prices are tracked
AI model token prices (input, output and cached input per 1M tokens) are refreshed automatically several times a day from live market feeds. Developer API and cloud plan prices are read from each vendor's official pricing page. Unusually large jumps are held for manual review before they go live.
Why do AI token prices change so often?
Labs reprice models after launches, hardware improvements and competition. Hosted open models also change as inference providers adjust their rates.
What does a cached input price mean?
Repeated prompt prefixes can be billed at a discounted cached rate, which is often 75–90% cheaper than fresh input tokens.