2026-08-24

Every AI model price, context and end-of-life change detected on 2026-08-24.

ChangeModelDetail
PRICE CHANGEdeepseek/deepseek-v4-flashoutput +14.6% · cache read +14.6% · input +14.6%
PRICE CHANGEdeepseek/deepseek-v4-flash-0731output -26.9% · cache read -17.8% · input -17.8%
PRICE CHANGEdeepseek/deepseek-v4-prooutput +31.5% · cache read +31.5% · input +31.5%
PRICE CHANGEminimax/minimax-m2.7output +25.0% · cache read +25.0% · input +25.0%
PRICE CHANGEmoonshotai/kimi-k2.7-codecache read +11.8%
PRICE CHANGEqwen/qwen3.5-397b-a17boutput +53.9% · cache read now charged · input +28.2%
PRICE CHANGEqwen/qwen3.6-27boutput -11.1% · cache read no longer charged · input -46.7%
NEW MODELmistralai/devstral-2512context 262,144
MODEL REMOVEDinclusionai/ling-2.6-1t
MODEL REMOVEDinclusionai/ling-2.6-flash
MODEL REMOVEDinclusionai/ring-2.6-1t
MODEL REMOVEDnvidia/nemotron-3-nano-30b-a3b:free
MODEL REMOVEDnvidia/nemotron-nano-12b-v2-vl:free
MODEL REMOVEDnvidia/nemotron-nano-9b-v2:free
OUTAGEElevated errors for multiple models
OUTAGEErrors logging into Claude.ai
RELEASE0.149.1
OUTAGEActions delays in starting runs
OUTAGEDegraded Git Operations over SSH
OUTAGEElevated errors on Fable 5 due to upstream provider
ANNOUNCEMENTAdvancing price-performance for developers with GPT‑5.6 in Kiro
PARAMETERS CHANGEDqwen/qwen-2.5-7b-instructremoved logit_bias
CONTEXT CHANGEDqwen/qwen3-235b-a22b-thinking-2507262144 -> 131072
PARAMETERS CHANGEDqwen/qwen3-235b-a22b-thinking-2507removed logit_bias, min_p
PARAMETERS CHANGEDqwen/qwen3.8-27badded min_p

Previous 2026-08-23 · Next 2026-08-25