2026-08-28

2026-08-28에 감지된 AI 모델 가격·컨텍스트·종료일 변경 전체입니다.

변경모델내용
종료 예정z-ai/glm-4.5v2026-12-31 종료 · D-124 남음
종료일 공지moonshotai/kimi-k2.5- -> 2026-08-31
종료일 변경~z-ai/glm-latest2098-12-31 -> -
가격 변경deepseek/deepseek-v3.2출력 +5.3% · 캐시 읽기 +3.5% · 입력 +3.5%
가격 변경deepseek/deepseek-v4-flash출력 +9.2% · 캐시 읽기 +9.2% · 입력 +9.2%
가격 변경deepseek/deepseek-v4-pro출력 -14.8% · 캐시 읽기 -14.8% · 입력 -14.8%
가격 변경deepseek/deepseek-v4-pro-0813출력 -41.2% · 캐시 읽기 -41.2% · overrides 과금 시작 · 입력 -41.2%
가격 변경google/gemini-3.7-flashaudio +100.0% · 출력 +100.0% · image +100.0% · input_audio_cache +100.0% · 캐시 읽기 +100.0% · 캐시 쓰기 +100.0% · internal_reasoning +100.0% · 입력 +100.0%
가격 변경moonshotai/kimi-k2.5출력 +33.3% · 캐시 읽기 +42.9% · 입력 +33.3%
가격 변경moonshotai/kimi-k2.7-code캐시 읽기 -5.3% · 입력 -1.5%
가격 변경nvidia/nemotron-3-nano-30b-a3b캐시 읽기 -16.7%
가격 변경nvidia/nemotron-3-ultra-550b-a55b출력 -38.9% · 캐시 읽기 -50.0% · 입력 -16.7%
가격 변경openai/gpt-4.1-nano캐시 읽기 +20.0%
가격 변경qwen/qwen3-235b-a22b-2507출력 -36.4% · 캐시 읽기 과금 시작 · 입력 -2.8%
가격 변경qwen/qwen3-vl-30b-a3b-instruct출력 +15.4% · 입력 +15.4%
가격 변경qwen/qwen3.5-122b-a10b출력 +15.4% · 입력 +11.5%
가격 변경qwen/qwen3.6-27b출력 +12.5% · 캐시 읽기 과금 시작 · 입력 +87.5%
가격 변경qwen/qwen3.6-35b-a3b출력 -10.0% · 입력 -28.6%
가격 변경qwen/qwen3.8-2.4t-a95b캐시 읽기 -20.0%
가격 변경z-ai/glm-4.6출력 -12.5% · 캐시 읽기 -20.0% · 입력 -14.0%
가격 변경~deepseek/deepseek-v4-flash-latest출력 +33.3%
가격 변경~google/gemini-flash-latestaudio +100.0% · 출력 +100.0% · image +100.0% · input_audio_cache +100.0% · 캐시 읽기 +100.0% · 캐시 쓰기 +100.0% · internal_reasoning +100.0% · 입력 +100.0%
가격 변경~z-ai/glm-latest입력 -10.7%
신규 모델inclusionai/ling-3.0-flash-fin:free컨텍스트 262,144
신규 모델mistralai/codestral-2508:batch컨텍스트 256,000
신규 모델mistralai/ministral-8b-2512:batch컨텍스트 262,144
신규 모델mistralai/mistral-large-2512:batch컨텍스트 262,144
신규 모델mistralai/mistral-medium-3-5:batch컨텍스트 262,144
신규 모델mistralai/mistral-medium-3.1:batch컨텍스트 131,072
신규 모델mistralai/mistral-small-2603:batch컨텍스트 262,144
신규 모델qwen/qwen3.8-2.4t-a95b:batch컨텍스트 1,010,000
신규 모델qwen/qwen3.8-flash컨텍스트 1,000,000
신규 모델tencent/hy4-preview컨텍스트 1,048,576
모델 삭제mistralai/ministral-8b
모델 삭제openai/gpt-5-codex:batch
모델 삭제thedrummer/rocinante-12b
당사자 발언Sam Altman @sama다음 모델 출시 때 파티를 또 해야 할 것 같아, 5.5 파티는 정말 재미있었어. 다음번 파티를 멋지게 만들어 줄 건 뭘까? || i think we should do another party for our next model release, the 5.5 party was a lot of fun. what would make the next one awesome?
당사자 발언Elon Musk @elonmusk여러분의 Grok @Bot을 다른 사람들과 공유하세요! https://t.co/hY9uqE2tsb || Share your Grok @Bot with others! https://t.co/hY9uqE2tsb
당사자 발언Tibo @thsottiauxChatGPT Work에 맡긴 가장 야심 찬 작업은 무엇인가요? || What is the most ambitious task you’ve given ChatGPT Work?
당사자 발언Tibo @thsottiaux@buildwithrajath 항상 개선 중입니다. 팀은 인프라 및 사용 최적화를 위해 열심히 작업하고 있습니다 || @buildwithrajath Always improving. Team working hard on infra and usage optimizations
당사자 발언Tibo @thsottiaux@TokenGremlin Luna는 세션에 제목을 붙이는 데 사용됩니다. 약간 “버터를 건네는” 상황 같은 거죠 || @TokenGremlin Luna is used to give the session a title. A bit of a pass “you pass the butter” situation
당사자 발언Anthropic @AnthropicAIClaude는 측정 가능한 정렬 불일치를 안정적으로 바로잡을 수 있습니다. 하지만 미묘하거나 드문 실패에는 벤치마크가 전혀 없을 수도 있습니다—따라서 모든 것은 올바른 것들을 측정하는 데 달려 있습니다. 다른 사람들이 이를 바탕으로 발전시킬 수 있도록 저희의 자동화된 정렬 연구 환경을 공개합니다. 전체 보고서: https://t.co/XpiMxgOonm || Claude can reliably fix measurable misalignment. But subtle or rare failures may have no benchmark at all—so everything hinges on measuring the right things. We're releasing our automated alignment re
당사자 발언Anthropic @AnthropicAI한 모델이 언젠가 자신보다 더 강력한 후속 모델들을 정렬할 수 있을까요? 첫 번째 테스트로, 저희는 Sonnet 5가 더 유능한 모델인 Opus 4.8의 초기 체크포인트를 사후 훈련하도록 했습니다. 이 모델은 저희의 전체 정렬 훈련을 거친 프로덕션 Opus 4.8의 점수에 근접하는 안전성 점수에 도달했습니다. https://t.co/FH3GezMLoZ || Could a model one day align its stronger successors? As a first test, we had Sonnet 5 post-train an early checkpoint of Opus 4.8, a more capable model. It reached safety scores approaching those of pr
당사자 발언Anthropic @AnthropicAI10개의 정렬 실패 전반에서 Claude는 역량을 저하시키지 않으면서 안전성 점수를 안정적으로 향상시켰습니다. Claude의 가장 뛰어난 방법들은 최적화 대상으로 삼지 않았던 벤치마크와 Petri 행동 감사, 그리고 최대 4.7x 더 큰 모델에도 일반화되었습니다. https://t.co/WD7FjlXXtc || Across 10 alignment failures, Claude reliably improved safety scores without degrading capabilities. Its best methods also generalized to benchmarks it hadn’t optimized on, to the Petri behavioral aud
당사자 발언Anthropic @AnthropicAIClaude는 기만이나 아첨 같은 흔한 정렬 불일치에 대한 안전성 벤치마크를 “힐 클라이밍”했습니다. 제약 조건은 하나였습니다: 일반 역량을 보존해야 한다는 것이었습니다. 그런 다음 그 방법들이 일반화되는지 확인하기 위해 Claude의 가장 뛰어난 방법들을 별도로 남겨 둔 벤치마크에서 테스트했습니다. || Claude “hill-climbed” safety benchmarks for common misalignments like deception or sycophancy, with one constraint: it had to preserve general capabilities. We then tested its best methods on held-out
당사자 발언Anthropic @AnthropicAI새로운 Fellows Research: Claude가 다른 AI들을 자율적으로 정렬할 수 있을까요? 저희는 소형 모델들의 정렬을 개선하도록 Claude에게 48 hours와 1 GPU를 주었습니다. Claude는 스스로 연구하고 방법을 제안한 다음, 모델들을 훈련하고 테스트했습니다. 놀라울 정도로 잘 작동했습니다. https://t.co/nhlCMgQl46 || New Fellows Research: Can Claude autonomously align other AIs? We gave Claude 48 hours and 1 GPU to improve the alignment of small models. It researched and proposed methods, then trained and tested t
OUTAGEElevated errors on Claude Code and Claude Cowork
ANNOUNCEMENTGemini Omni 1.1 Flash lets you build with more control
ANNOUNCEMENTPiloting the world's first double-blind AI evaluations
RELEASEv1.2.0
RELEASEv2.1.251
RELEASE0.150.1
RELEASElangchain-core==1.6.1
RELEASEv0.33.1
RELEASEv3.5.0
ANNOUNCEMENTEnterprise-managed settings now support autoUpdate for plugin marketplaces
ANNOUNCEMENTGlobal model policy generally available
ANNOUNCEMENTActions retention will cover checks, workflow runs, and statuses
ANNOUNCEMENTClose all open contributions authored by a blocked user
ANNOUNCEMENTCopilot code review: Resolution reasons and expanded capabilities
ANNOUNCEMENTGitHub Classroom deprecated
ANNOUNCEMENTBetter label management on issues is generally available
ANNOUNCEMENTUpcoming changes to GitHub Copilot policies and billing
ANNOUNCEMENTGitHub Copilot app for Beginners: Automate Dependabot pull request triage
ANNOUNCEMENTOpenClaw went viral. Meet the maintainers building and securing it.
OUTAGEDisruption with GitHub Billing
OUTAGEIncident with Actions and Pull Requests
OUTAGEIncident with Copilot AI Model Providers
WEIGHT_RELEASEgoogle/timesfm-3.0-pytorch
ANNOUNCEMENTExpanding OpenAI’s presence in Brazil
ANNOUNCEMENTThe Hugging Face incident and the road ahead
ANNOUNCEMENTIntroducing Intelligence Age
ANNOUNCEMENTSupporting Thailand’s next generation of AI startups
ANNOUNCEMENTBetter answers, broader thinking: What students gain from ChatGPT and critical-thinking training
OUTAGEUsers may experience an increase in error rates in Workspace Agents and ChatGPT Work on Web and Mobile
변형 표기 삭제moonshotai/kimi-k2.7-code:batch':batch' 표기가 더 이상 목록에 없습니다 - moonshotai/kimi-k2.7-code 자체는 그대로입니다
변형 표기 삭제openai/gpt-3.5-turbo:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-3.5-turbo 자체는 그대로입니다
변형 표기 삭제openai/gpt-4-turbo:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-4-turbo 자체는 그대로입니다
변형 표기 삭제openai/gpt-4.1-mini:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-4.1-mini 자체는 그대로입니다
변형 표기 삭제openai/gpt-4.1-nano:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-4.1-nano 자체는 그대로입니다
변형 표기 삭제openai/gpt-4.1:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-4.1 자체는 그대로입니다
변형 표기 삭제openai/gpt-4o-mini:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-4o-mini 자체는 그대로입니다
변형 표기 삭제openai/gpt-4o:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-4o 자체는 그대로입니다
변형 표기 삭제openai/gpt-5-mini:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5-mini 자체는 그대로입니다
변형 표기 삭제openai/gpt-5-nano:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5-nano 자체는 그대로입니다
변형 표기 삭제openai/gpt-5-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.1:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.1 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.2-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.2-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.2:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.2 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.4-mini:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.4-mini 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.4-nano:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.4-nano 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.4-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.4-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.4:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.4 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.5-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.5-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.5:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.5 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.6-luna-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.6-luna-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.6-luna:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.6-luna 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.6-sol-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.6-sol-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.6-sol:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.6-sol 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.6-terra-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.6-terra-pro 자체는 그대로입니다
변형 표기 삭제openai/gpt-5.6-terra:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5.6-terra 자체는 그대로입니다
변형 표기 삭제openai/gpt-5:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/gpt-5 자체는 그대로입니다
변형 표기 삭제openai/o1-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o1-pro 자체는 그대로입니다
변형 표기 삭제openai/o1:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o1 자체는 그대로입니다
변형 표기 삭제openai/o3-mini-high:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o3-mini-high 자체는 그대로입니다
변형 표기 삭제openai/o3-mini:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o3-mini 자체는 그대로입니다
변형 표기 삭제openai/o3-pro:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o3-pro 자체는 그대로입니다
변형 표기 삭제openai/o3:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o3 자체는 그대로입니다
변형 표기 삭제openai/o4-mini-high:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o4-mini-high 자체는 그대로입니다
변형 표기 삭제openai/o4-mini:batch':batch' 표기가 더 이상 목록에 없습니다 - openai/o4-mini 자체는 그대로입니다
변형 표기 삭제z-ai/glm-5.2:batch':batch' 표기가 더 이상 목록에 없습니다 - z-ai/glm-5.2 자체는 그대로입니다
컨텍스트 변경google/gemma-3-27b-it262144 -> 131072
컨텍스트 변경kwaipilot/kat-coder-pro-v2.5256000 -> 262144
파라미터 변경kwaipilot/kat-coder-pro-v2.5추가 frequency_penalty, logit_bias, min_p, repetition_penalty, seed, top_k
컨텍스트 변경mistralai/voxtral-small-24b-250732000 -> 32768
컨텍스트 변경nvidia/nemotron-3-ultra-550b-a55b512288 -> 262144
파라미터 변경qwen/qwen3-next-80b-a3b-thinking삭제 logit_bias
컨텍스트 변경z-ai/glm-5.31048576 -> 1310720
파라미터 변경z-ai/glm-5.3추가 frequency_penalty, logit_bias, logprobs, min_p, presence_penalty, repetition_penalty, seed, stop, structured_outputs, top_logprobs
컨텍스트 변경~z-ai/glm-latest1048576 -> 1310720
파라미터 변경~z-ai/glm-latest추가 frequency_penalty, logit_bias, logprobs, min_p, presence_penalty, repetition_penalty, seed, stop, structured_outputs, top_logprobs

이전 2026-08-26 · 다음 2026-08-29