AI Calculator Pro
Weekly recap

What changed in LLM pricing — Week of 17–23 August 2026

In the week of 17–23 August 2026, we recorded 31 price changes (29 cuts, 2 increases) and 15 new models across the LLM providers we track. The biggest cut was Qwen3.8 27B cached input, down 88% ($0.080 → $0.010 per 1M tokens).

The median cut this week was 20%. Figures are from our daily re-check of official provider pricing; see the live tracker for the full history.

31
Price changes
29
Price drops
2
Price increases
15
New models

Price cuts in Week of 17–23 August 2026

Models that got cheaper, biggest cut first. Click a column to sort.

Qwen3.8 27BCached input$0.080/1M$0.010/1M-88%17 August 2026Alibaba
Gemini 3.6 FlashInput$1.50/1M$0.750/1M-50%21 August 2026Google
Gemini 3.6 FlashOutput$7.50/1M$3.75/1M-50%21 August 2026Google
Gemini 3.6 FlashCached input$0.150/1M$0.075/1M-50%21 August 2026Google
GPT 5.6 Sol ProCache write$6.25/1M$3.13/1M-50%18 August 2026OpenAI
GPT 5.6 Sol ProCached input$0.500/1M$0.250/1M-50%18 August 2026OpenAI
GPT 5.6 Sol ProOutput$30.00/1M$15.00/1M-50%18 August 2026OpenAI
GPT 5.6 Sol ProInput$5.00/1M$2.50/1M-50%18 August 2026OpenAI

Price increases in Week of 17–23 August 2026

Nemotron 3 Nano OmniCached input$0.006/1M$0.054/1M+800%21 August 2026Nvidia
Gemma 4 31B ITCached input$0.010/1M$0.040/1M+300%22 August 2026Google

Models launched in Week of 17–23 August 2026

Cite this recap

Monthly recaps are generated from our open price-change dataset (CC BY 4.0). See the recap JSON, the full change dataset, or the RSS feed.