2026-09-03

2026-09-03에 감지된 AI 모델 가격·컨텍스트·종료일 변경 전체입니다.

변경모델내용
종료 예정nex-agi/nex-n2-mini2026-09-08 종료 · D-5 남음
종료 예정nex-agi/nex-n2-pro2026-09-08 종료 · D-5 남음
종료 예정dots-studio/dots-3-note-preview:free2026-09-30 종료 · D-27 남음
종료 예정z-ai/glm-4.52026-12-31 종료 · D-119 남음
종료 예정z-ai/glm-4.5v2026-12-31 종료 · D-119 남음
종료일 변경z-ai/glm-5.32098-12-31 -> -
종료일 변경~z-ai/glm-flash-latest2098-12-31 -> -
가격 변경deepseek/deepseek-chat출력 -13.5% · 입력 +24.3%
가격 변경deepseek/deepseek-chat-v3.1출력 +73.7% · 캐시 읽기 +323.1% · 입력 +120.0%
가격 변경deepseek/deepseek-v4-flash출력 -1.9% · 캐시 읽기 -1.9% · 입력 -1.9%
가격 변경deepseek/deepseek-v4-pro출력 +0.3% · 캐시 읽기 +0.3% · 입력 +0.3%
가격 변경meta-llama/llama-3.3-70b-instruct출력 -54.9% · 캐시 읽기 과금 중단 · 입력 -85.9%
가격 변경meta/muse-glimmer-30b출력 -8.3%
가격 변경nvidia/nemotron-3-nano-30b-a3b캐시 읽기 +20.0%
가격 변경nvidia/nemotron-3-ultra-550b-a55b출력 -23.2% · 캐시 읽기 -36.0% · 입력 -4.0%
가격 변경qwen/qwen2.5-vl-72b-instruct출력 +33.3% · 캐시 읽기 과금 시작 · 입력 +220.0%
가격 변경qwen/qwen3.5-397b-a17b출력 +49.6% · 캐시 읽기 과금 시작 · 입력 +41.0%
가격 변경qwen/qwen3.8-2.4t-a95b캐시 읽기 +25.0%
가격 변경qwen/qwen3.8-27b출력 +17.6% · 캐시 쓰기 과금 중단 · 입력 -1.2%
가격 변경z-ai/glm-4.6출력 +25.7% · 캐시 읽기 +37.5% · 입력 +27.9%
가격 변경z-ai/glm-5.3캐시 읽기 -46.1%
가격 변경~moonshotai/kimi-latest출력 +9.8% · 캐시 읽기 +13.3% · 입력 -2.0%
가격 변경~z-ai/glm-latest출력 +3.1% · 캐시 읽기 +88.6% · 입력 -0.2%
신규 모델inclusionai/ling-3.0-flash-fin컨텍스트 262,144
신규 모델meta/muse-spark-1.3컨텍스트 1,048,576
신규 모델meta/muse-spark-1.3-contributor컨텍스트 1,048,576
모델 삭제nvidia/nemotron-3-ultra-550b-a55b:batch
당사자 발언Tibo @thsottiaux@TokenGremlin 내가 그걸 대략 그려볼게 || @TokenGremlin Let me sketch that out
당사자 발언Tibo @thsottiaux@prashantmital Codex한테 물어봤어? || @prashantmital Did you ask Codex?
OUTAGEElevated errors for Claude Sonnet 5
OUTAGEElevated errors for multiple models
OUTAGEElevated errors for Claude Sonnet 5
ANNOUNCEMENTIntroducing WeatherNext 3, our most advanced and accurate global weather AI model
RELEASEv2.1.259
RELEASE0.153.0
RELEASEv0.33.3
ANNOUNCEMENTContent exclusions generally available in Copilot app and CLI
ANNOUNCEMENTEnterprise-managed settings support any default model
ANNOUNCEMENTCodeQL 2.26.4 improves GitHub actions security detections
ANNOUNCEMENTReopening Copilot Business and Enterprise signups
ANNOUNCEMENTDecoding the new AI lingo: Loops, harnesses, squads, hill climbing… oh my!
ANNOUNCEMENTGitHub Copilot app for Beginners: Run several agents at once
OUTAGEIncident with Grok Copilot AI Model Provider
ANNOUNCEMENTNeoMME: an efficient Multimodal-native and Multilingual Encoder
ANNOUNCEMENTGive Your Coding Agents a Memory You Own
ANNOUNCEMENTFine-tuning a 350M Model for Better Structured Outputs in 100 GRPO Steps
ANNOUNCEMENTTraining a coding model to paint watercolours with TRL and OpenEnv
ANNOUNCEMENTATV Big Air Tour turned 3 days of work into 3 hours with ChatGPT
OUTAGEChatGPT Work Mode High Error Rates
OUTAGEElevated errors across ChatGPT and Codex
파라미터 변경deepseek/deepseek-v4-flash-vision-exp추가 min_p
파라미터 변경ibm-granite/granite-4.2-8b추가 logit_bias, min_p

이전 2026-09-02