Weekly recap
What changed in LLM pricing — Week of 21–27 September 2026
In the week of 21–27 September 2026, we recorded 6 price changes (3 cuts, 3 increases) and 2 new models across the LLM providers we track. The biggest cut was deepseek-v4-1-flash output, down 30% ($0.400 → $0.280 per 1M tokens).
The median cut this week was 13%. Figures are from our daily re-check of official provider pricing; see the live tracker for the full history.
6
Price changes
3
Price drops
3
Price increases
2
New models
Price cuts in Week of 21–27 September 2026
Models that got cheaper, biggest cut first. Click a column to sort.
| deepseek-v4-1-flash | Output | $0.400/1M | $0.280/1M | -30% | 21 September 2026 | DeepSeek |
| Qwen3.8 Omni Flash | Input | $0.150/1M | $0.130/1M | -13% | 21 September 2026 | Alibaba |
| Qwen3.8 Omni Flash | Output | $0.470/1M | $0.430/1M | -9% | 21 September 2026 | Alibaba |
Price increases in Week of 21–27 September 2026
| deepseek-v4-1-flash | Input | $0.100/1M | $0.140/1M | +40% | 21 September 2026 | DeepSeek |
| Nemotron 3.5 Lightning 30B | Input | $0.040/1M | $0.050/1M | +25% | 21 September 2026 | Nvidia |
| Nemotron 3.5 Lightning 30B | Output | $0.180/1M | $0.200/1M | +11% | 21 September 2026 | Nvidia |
Models launched in Week of 21–27 September 2026
- deepseek-v4-1-flash· DeepSeek$0.140/1M · 21 September 2026
- Qwen3.8 Omni Flash· Alibaba$0.130/1M · 21 September 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.