2026-08-25

Every AI model price, context and end-of-life change detected on 2026-08-25.

ChangeModelDetail
END OF LIFEmoonshotai/kimi-k2.5shuts down 2026-08-31 · 6 days left
END OF LIFEdots-studio/dots-3-note-preview:freeshuts down 2026-09-30 · 36 days left
END OF LIFEz-ai/glm-4.5shuts down 2026-12-31 · 128 days left
END OF LIFE MOVEDz-ai/glm-4.5v2026-12-31 -> -
PRICE CHANGEdeepseek/deepseek-v4-flashoutput +47.5% · cache read +47.5% · input +47.5%
PRICE CHANGEdeepseek/deepseek-v4-flash-0731output -39.2% · cache read -39.2% · input -39.2%
PRICE CHANGEdeepseek/deepseek-v4-flash-vision-expoverrides now charged
PRICE CHANGEdeepseek/deepseek-v4-prooutput +9.7% · cache read +9.7% · input +9.7%
PRICE CHANGEqwen/qwen3.8-27boutput -15.0%
PRICE CHANGEthinkingmachines/inklingcache read -5.9% · input -5.0%
PRICE CHANGEz-ai/glm-5.1output +30.4% · cache read +30.4% · input +30.4%
PRICE CHANGEz-ai/glm-5.2output +23.2% · cache read +14.4% · input +23.2%
PRICE CHANGE~deepseek/deepseek-v4-flash-latestoutput +25.0% · input -12.5%
PRICE CHANGE~moonshotai/kimi-latestoutput +7.7% · input +7.7%
NEW MODELminimax/minimax-m2.7:freecontext 196,608
NEW MODELminimax/minimax-m3:freecontext 1,048,576
MODEL REMOVEDqwen/qwen-plus-2025-07-28:thinking
FROM THE SOURCEOpenAI @OpenAIJalapeño means faster ChatGPT responses, more responsive Codex sessions and agents, and reliable access as demand continues to grow. https://t.co/QgBKRa3Xz6
FROM THE SOURCEOpenAI @OpenAISince announcing Jalapeño, our first custom inference chip, we’ve been testing it and the system around it. The results show a major advance: more intelligence from every watt and faster responses, delivering both higher throughput and lowe
OUTAGEIssues logging into Claude.ai
RELEASEv2.1.245
RELEASEv0.3.0
ANNOUNCEMENTPush rules in rulesets now support path exceptions
ANNOUNCEMENTYour alt text passes automated checks. That doesn’t mean it’s any good.
ANNOUNCEMENTQuantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision original
ANNOUNCEMENTWire It, Run It, Deploy It: AI Workflows in Gradio
ANNOUNCEMENTGranite 4.2 LLMs: How They're Built
ANNOUNCEMENTHow Hugging Face Inference Endpoints, Jobs, and Buckets Power Search on Papers with Code
ANNOUNCEMENTDisrupting a new covert influence campaign from Russia
ANNOUNCEMENTIntroducing the Admin plugin for ChatGPT Work and Codex
ANNOUNCEMENTJalapeño’s first results show industry-leading speed and efficiency in AI inference
ANNOUNCEMENTThe full stack behind abundant intelligence
CONTEXT CHANGEDthinkingmachines/inkling-small:free262144 -> 1048576
CONTEXT CHANGEDthinkingmachines/inkling:free262144 -> 1048576
PARAMETERS CHANGEDx-ai/grok-4.6added stop, top_k
PARAMETERS CHANGED~x-ai/grok-latestadded stop, top_k

Previous 2026-08-24