2026-08-18에 감지된 AI 모델 가격·컨텍스트·종료일 변경 전체입니다.
| 변경 | 모델 | 내용 |
|---|---|---|
| 종료 예정 | nvidia/nemotron-nano-12b-v2-vl:free | 2026-08-24 종료 · D-6 남음 |
| 종료 예정 | z-ai/glm-4.5 | 2026-12-31 종료 · D-135 남음 |
| 종료일 공지 | nvidia/nemotron-nano-12b-v2-vl:free | - -> 2026-08-24 |
| 가격 변경 | deepseek/deepseek-v4-flash | 출력 +3.6% · 캐시 읽기 +3.6% · 입력 +3.6% |
| 가격 변경 | ~deepseek/deepseek-v4-flash-latest | 출력 +0.2% · 캐시 읽기 +0.2% · 입력 +0.2% |
| 당사자 발언 | Sam Altman @sama | We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Model progress is now extremely rapid, and we always said we |
| 당사자 발언 | Tibo @thsottiaux | Gimme, gimme, gimme Codex after midnight Won’t somebody make these failing tests all go away? Gimme, gimme, gimme Codex after midnight Ship it through the darkness by the start of the day |
| 당사자 발언 | Tibo @thsottiaux | What is an obvious thing that we should do with Codex, API or our models that we should just do but haven't yet? What is 100% within reach, but we just seem to be missing? |
| 당사자 발언 | OpenAI @OpenAI | As models become more capable, the risks associated with developing and testing them internally also grow. We temporarily paused reinforcement learning (RL) training on our latest models intended for deployment for two weeks while we harden |
| 당사자 발언 | Google AI Developers @googleaidevs | Turn the web into an interactive visual encyclopedia with Nano Banana and Gemini Omni. 🌐✨ We built a Chrome Extension in @Antigravity with Gemini 3.7 Flash that lets you highlight any text to quickly generate rich, visual definitions while |
| 당사자 발언 | Google AI Developers @googleaidevs | We fed Gemini 3.7 Flash hundreds of PDFs from a Victorian book of botanical illustrations and tasked it with extracting and classifying these historical plants using the modern APG IV classification system. 🌳 💻 Watch how it brings its under |
| OUTAGE | Degraded performance for multiple models | |
| RELEASE | v2.1.234 | |
| RELEASE | langchain-core==1.5.6 | |
| RELEASE | b10488 | |
| OUTAGE | Intermittent failures in runner group and runner-related permissions pages | |
| OUTAGE | Incident with Actions | |
| ANNOUNCEMENT | Same Cluster, 33 Points More Utilization: What Changed Was the Order | |
| ANNOUNCEMENT | How Much Memory Does Your Agent Actually Need? | |
| ANNOUNCEMENT | Multi-Vector (Late Interaction) Embedding Models with Sentence Transformers | |
| ANNOUNCEMENT | Asana cleared 5 years of engineering work in 2 weeks with Codex | |
| ANNOUNCEMENT | Introducing ChatGPT for Teens: Built for learning, backed by protections | |
| ANNOUNCEMENT | Pacing model development in an era of cyber-critical capabilities | |
| ANNOUNCEMENT | Partnering with CodeAI to prepare the first AI generation | |
| 파라미터 변경 | inclusionai/ling-3.0-flash | 추가 response_format, structured_outputs |
| 파라미터 변경 | x-ai/grok-4.3 | 삭제 frequency_penalty, presence_penalty, stop |
| 파라미터 변경 | x-ai/grok-4.5 | 삭제 frequency_penalty, presence_penalty, stop |
| 파라미터 변경 | x-ai/grok-4.6 | 삭제 frequency_penalty, presence_penalty, stop |
| 파라미터 변경 | x-ai/grok-build-0.1 | 삭제 frequency_penalty, presence_penalty, stop |
| 컨텍스트 변경 | z-ai/glm-5.2:free | 128000 -> 256000 |
| 파라미터 변경 | z-ai/glm-5.2:free | 추가 response_format, structured_outputs, tool_choice, tools |
| 파라미터 변경 | ~x-ai/grok-latest | 삭제 frequency_penalty, presence_penalty, stop |