Skip to content

Commit 7c44782

Browse files
authored
Merge pull request #40 from NHLOCAL/update
December update
2 parents 7fea316 + bd1dfa2 commit 7c44782

File tree

3 files changed

+63
-9
lines changed

3 files changed

+63
-9
lines changed

_data/timeline.md

Lines changed: 17 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -135,7 +135,7 @@
135135
- OPENAI has released two nextgeneration AI models to its subscribers: **o1 preview** and **o1 mini**. These models show a significant improvement in performance, particularly in tasks requiring reasoning, including coding, mathematics, GPQA, and more. (*special*)
136136
- Chinese company Alibaba releases the **Qwen 2.5** model in various sizes, ranging from 0.5B to 72B. The models demonstrate capabilities comparable to much larger models.
137137
- The video generation model **KLING 1.5** has been released.
138-
- **OpenAI** launches the **advanced voice mode** of GPT4o for all subscribers.
138+
- **OpenAI** launches the **advanced voice mode** of GPT 4o for all subscribers.
139139
- **Meta** releases **Llama 3.2** in sizes 1B, 3B, 11B and 90B, featuring image recognition capabilities for the first time.
140140
- **Google** has rolled out new model updates ready for deployment, **Gemini Pro 1.5 002** and **Gemini Flash 1.5 002**, showcasing significantly improved longcontext processing.
141141
- **Kyutai** releases two opensource versions of its voicetovoice model, **Moshi**.
@@ -169,7 +169,7 @@
169169

170170
## December
171171
- Amazon introduced a new series of models called **NOVA**, designed for text, image, and video processing.
172-
- OpenAI released **SORA**, a video generation model, along with the full version of **O1** and **O1 Pro** for advanced subscribers. Additionally, the company launched a live video mode for **GPT4o**. (*special*)
172+
- OpenAI released **SORA**, a video generation model, along with the full version of **o1** and **o1 Pro** for advanced subscribers. Additionally, the company launched a live video mode for **GPT 4o**. (*special*)
173173
- Google unveiled the experimental model **Gemini-Exp-1206**, which ranked first in the chatbot leaderboard.
174174
- Google launched **Gemini 2.0 Flash** in beta. This model leads benchmarks and outperforms the previous version, **Gemini Pro 1.5**. Additionally, Google introduced live speech and video mode and announced built-in image generation capabilities within the model. (*special*)
175175
- Google revealed **Gemini-2.0-Flash-Thinking**, a thinking model based on **Gemini 2.0 Flash**, which secured second place in the chatbot leaderboard. (*special*)
@@ -182,7 +182,7 @@
182182
- Meta introduced **Apollo**, a video generation model available in three different sizes.
183183
- Deepseek open-sourced **Deepseek V3**, a model with 671B parameters that surpasses closed-source SOTA models across several benchmarks. (*special*)
184184
- Alibaba unveiled **QVQ-72B-Preview**, a cutting-edge thinking model capable of analyzing images, featuring SOTA-level performance. (*special*)
185-
- OpenAI announced **O3**, a groundbreaking AI model achieving 87.5% in the **ARC-AGI** benchmark, 25.2% in the **Frontier Math Benchmark** (compared to under 2% in previous models), and 87.7% in Ph.D.-level science questions. A cost-effective version, **O3 Mini**, is expected in January 2025, with performance similar to **O1**, alongside improved speed and efficiency. (*special*)
185+
- OpenAI announced **o3**, a groundbreaking AI model achieving 87.5% in the **ARC-AGI** benchmark, 25.2% in the **Frontier Math Benchmark** (compared to under 2% in previous models), and 87.7% in Ph.D.-level science questions. A cost-effective version, **o3 Mini**, is expected in January 2025, with performance similar to **o1**, alongside improved speed and efficiency. (*special*)
186186
- The video generation model **Kling 1.6** was released, offering significant performance enhancements.
187187

188188

@@ -221,7 +221,7 @@
221221
- Google launches **Gemini 2.5 Flash**, with a dynamic reasoning mode that allows tuning the reasoning level or disabling it as needed.
222222
- Amazon introduces **Nova Act**, a new framework for building multi-step autonomous agents.
223223
- OpenAI releases **GPT-4.1** in three sizes, with a context window of 1 million tokens.
224-
- OpenAI introduces **O3 full** and **O4 mini**, highly advanced models for reasoning, math, and coding.
224+
- OpenAI introduces **o3 full** and **o4 mini**, highly advanced models for reasoning, math, and coding.
225225
- Midjourney launches **v7**, with higher image quality and more precise control over style.
226226
- A series of video model updates - **Veo 2.0** (Google), **Runway Gen-4**, **Vidu Q1**, and **Kling 2.0** – a leap forward in high-quality video generation, with improvements in response times, realism, and style.
227227
- Alibaba releases **Qwen 3** as open source, in various sizes, with very impressive capabilities for their size. (*special*)
@@ -300,3 +300,16 @@
300300
- DeepSeek released **DeepSeekMath-V2** as open source, achieving gold-medal performance in math olympiads. (*special*)
301301
- Microsoft open-sourced **Fara-7B**, a small model optimized for browser agents and computer control.
302302
- **Poetiq** shatters the **ARC-AGI-2** benchmark with a score of over 60%, surpassing the human average.
303+
304+
305+
## December 2025
306+
- Mistral AI launches the **Mistral 3** family (Large & Ministral) alongside **Mistral OCR 3** and the **Devstral 2** coding series, reinforcing its open-weight leadership with advanced agentic workflows and Vibe CLI integration.
307+
- OpenAI releases **GPT-5.2**, featuring the autonomous **Codex** agent for complex engineering tasks, and **GPT-Image 1.5**, which claims the #1 spot on vision benchmarks, outperforming Nano Banana Pro.
308+
- Google introduces **Gemini 3.0 Flash**, setting a new standard for price-performance, and deploys **Deep Research**, an autonomous agent capable of multi-step synthesis, alongside **Gemini 2.5 Flash Audio**. (*special*)
309+
- Amazon unveils the **Nova 2** series, highlighted by **Nova 2 Sonic**, a native speech-to-speech model delivering ultra-low latency and natural conversation flow.
310+
- Runway releases **Gen-4.5**, a video generation model that rises to the top of industry leaderboards for motion consistency and prompt adherence.
311+
- xAI launches the **Grok Voice Agent API**, enabling native, real-time bidirectional audio streaming for developers.
312+
- Zhipu AI releases **GLM-4.7**, an open-weights model that reaches the top of global coding and reasoning leaderboards.
313+
- Alibaba open-sources **Z-Image-Turbo**, a highly efficient 6B model, and releases **Qwen-Image-2512**, which specializes in high-fidelity typography and complex visual compositions.
314+
- MiniMax releases **MiniMax-M2.1**, a 200k-context MoE model that rises to the top of web development and coding leaderboards, establishing itself as a leading open model for developers.
315+
- A specialized system by **Poetiq**, powered by GPT-5.2, reportedly solves the **ARC-2** benchmark, marking a major breakthrough in abstract reasoning. (*special*)

_data/timeline.yml

Lines changed: 18 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -149,7 +149,7 @@
149149
special: true
150150
- text: Chinese company Alibaba releases the <b>Qwen 2.5</b> model in various sizes, ranging from 0.5B to 72B. The models demonstrate capabilities comparable to much larger models.
151151
- text: The video generation model <b>KLING 1.5</b> has been released.
152-
- text: <b>OpenAI</b> launches the <b>advanced voice mode</b> of GPT4o for all subscribers.
152+
- text: <b>OpenAI</b> launches the <b>advanced voice mode</b> of GPT 4o for all subscribers.
153153
- text: <b>Meta</b> releases <b>Llama 3.2</b> in sizes 1B, 3B, 11B and 90B, featuring image recognition capabilities for the first time.
154154
- text: <b>Google</b> has rolled out new model updates ready for deployment, <b>Gemini Pro 1.5 002</b> and <b>Gemini Flash 1.5 002</b>, showcasing significantly improved longcontext processing.
155155
- text: <b>Kyutai</b> releases two opensource versions of its voicetovoice model, <b>Moshi</b>.
@@ -183,7 +183,7 @@
183183
- date: December
184184
info:
185185
- text: Amazon introduced a new series of models called <b>NOVA</b>, designed for text, image, and video processing.
186-
- text: OpenAI released <b>SORA</b>, a video generation model, along with the full version of <b>O1</b> and <b>O1 Pro</b> for advanced subscribers. Additionally, the company launched a live video mode for <b>GPT4o</b>.
186+
- text: OpenAI released <b>SORA</b>, a video generation model, along with the full version of <b>o1</b> and <b>o1 Pro</b> for advanced subscribers. Additionally, the company launched a live video mode for <b>GPT 4o</b>.
187187
special: true
188188
- text: Google unveiled the experimental model <b>Gemini-Exp-1206</b>, which ranked first in the chatbot leaderboard.
189189
- text: Google launched <b>Gemini 2.0 Flash</b> in beta. This model leads benchmarks and outperforms the previous version, <b>Gemini Pro 1.5</b>. Additionally, Google introduced live speech and video mode and announced built-in image generation capabilities within the model.
@@ -202,7 +202,7 @@
202202
special: true
203203
- text: Alibaba unveiled <b>QVQ-72B-Preview</b>, a cutting-edge thinking model capable of analyzing images, featuring SOTA-level performance.
204204
special: true
205-
- text: OpenAI announced <b>O3</b>, a groundbreaking AI model achieving 87.5% in the <b>ARC-AGI</b> benchmark, 25.2% in the <b>Frontier Math Benchmark</b> (compared to under 2% in previous models), and 87.7% in Ph.D.-level science questions. A cost-effective version, <b>O3 Mini</b>, is expected in January 2025, with performance similar to <b>O1</b>, alongside improved speed and efficiency.
205+
- text: OpenAI announced <b>o3</b>, a groundbreaking AI model achieving 87.5% in the <b>ARC-AGI</b> benchmark, 25.2% in the <b>Frontier Math Benchmark</b> (compared to under 2% in previous models), and 87.7% in Ph.D.-level science questions. A cost-effective version, <b>o3 Mini</b>, is expected in January 2025, with performance similar to <b>o1</b>, alongside improved speed and efficiency.
206206
special: true
207207
- text: The video generation model <b>Kling 1.6</b> was released, offering significant performance enhancements.
208208
- year: 2025
@@ -253,7 +253,7 @@
253253
- text: Google launches <b>Gemini 2.5 Flash</b>, with a dynamic reasoning mode that allows tuning the reasoning level or disabling it as needed.
254254
- text: Amazon introduces <b>Nova Act</b>, a new framework for building multi-step autonomous agents.
255255
- text: OpenAI releases <b>GPT-4.1</b> in three sizes, with a context window of 1 million tokens.
256-
- text: OpenAI introduces <b>O3 full</b> and <b>O4 mini</b>, highly advanced models for reasoning, math, and coding.
256+
- text: OpenAI introduces <b>o3 full</b> and <b>o4 mini</b>, highly advanced models for reasoning, math, and coding.
257257
- text: Midjourney launches <b>v7</b>, with higher image quality and more precise control over style.
258258
- text: A series of video model updates - <b>Veo 2.0</b> (Google), <b>Runway Gen-4</b>, <b>Vidu Q1</b>, and <b>Kling 2.0</b> – a leap forward in high-quality video generation, with improvements in response times, realism, and style.
259259
- text: Alibaba releases <b>Qwen 3</b> as open source, in various sizes, with very impressive capabilities for their size.
@@ -345,3 +345,17 @@
345345
special: true
346346
- text: Microsoft open-sourced <b>Fara-7B</b>, a small model optimized for browser agents and computer control.
347347
- text: <b>Poetiq</b> shatters the <b>ARC-AGI-2</b> benchmark with a score of over 60%, surpassing the human average.
348+
- date: December 2025
349+
info:
350+
- text: Mistral AI launches the <b>Mistral 3</b> family (Large & Ministral) alongside <b>Mistral OCR 3</b> and the <b>Devstral 2</b> coding series, reinforcing its open-weight leadership with advanced agentic workflows and Vibe CLI integration.
351+
- text: 'OpenAI releases <b>GPT-5.2</b>, featuring the autonomous <b>Codex</b> agent for complex engineering tasks, and <b>GPT-Image 1.5</b>, which claims the #1 spot on vision benchmarks, outperforming Nano Banana Pro.'
352+
- text: Google introduces <b>Gemini 3.0 Flash</b>, setting a new standard for price-performance, and deploys <b>Deep Research</b>, an autonomous agent capable of multi-step synthesis, alongside <b>Gemini 2.5 Flash Audio</b>.
353+
special: true
354+
- text: Amazon unveils the <b>Nova 2</b> series, highlighted by <b>Nova 2 Sonic</b>, a native speech-to-speech model delivering ultra-low latency and natural conversation flow.
355+
- text: Runway releases <b>Gen-4.5</b>, a video generation model that rises to the top of industry leaderboards for motion consistency and prompt adherence.
356+
- text: xAI launches the <b>Grok Voice Agent API</b>, enabling native, real-time bidirectional audio streaming for developers.
357+
- text: Zhipu AI releases <b>GLM-4.7</b>, an open-weights model that reaches the top of global coding and reasoning leaderboards.
358+
- text: Alibaba open-sources <b>Z-Image-Turbo</b>, a highly efficient 6B model, and releases <b>Qwen-Image-2512</b>, which specializes in high-fidelity typography and complex visual compositions.
359+
- text: MiniMax releases <b>MiniMax-M2.1</b>, a 200k-context MoE model that rises to the top of web development and coding leaderboards, establishing itself as a leading open model for developers.
360+
- text: A specialized system by <b>Poetiq</b>, powered by GPT-5.2, reportedly solves the <b>ARC-2</b> benchmark, marking a major breakthrough in abstract reasoning.
361+
special: true

notes.md

Lines changed: 28 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -86,4 +86,31 @@
8686
- חברת Anthropic הכריזה על **Claude Opus 4.5**, המציע יכולות קוד וסוכנים מובילות במחיר מופחת משמעותית.
8787
- מעבדות Black Forest Labs שחררו את **FLUX 2**, מודל תמונה בקוד פתוח עם ביצועים גבוהים.
8888
- חברת DeepSeek שחררה את **DeepSeekMath-V2** כקוד פתוח, שהשיג ביצועי מדליית זהב באולימפיאדות מתמטיקה.
89-
- מיקרוסופט שחררה כקוד פתוח את **Fara-7B**, מודל קטן המותאם לסוכני דפדפן ושליטה במחשב.
89+
- מיקרוסופט שחררה כקוד פתוח את **Fara-7B**, מודל קטן המותאם לסוכני דפדפן ושליטה במחשב.
90+
91+
92+
## December 2025
93+
- Mistral AI launches the **Mistral 3** family (Large & Ministral) alongside **Mistral OCR 3** and the **Devstral 2** coding series, reinforcing its open-weight leadership with advanced agentic workflows and Vibe CLI integration.
94+
- OpenAI releases **GPT-5.2**, featuring the autonomous **Codex** agent for complex engineering tasks, and **GPT-Image 1.5**, which claims the #1 spot on vision benchmarks, outperforming Nano Banana Pro.
95+
- Google introduces **Gemini 3.0 Flash**, setting a new standard for price-performance, and deploys **Deep Research**, an autonomous agent capable of multi-step synthesis, alongside **Gemini 2.5 Flash Audio**.
96+
- Amazon unveils the **Nova 2** series, highlighted by **Nova 2 Sonic**, a native speech-to-speech model delivering ultra-low latency and natural conversation flow.
97+
- Runway releases **Gen-4.5**, a video generation model that rises to the top of industry leaderboards for motion consistency and prompt adherence.
98+
- xAI launches the **Grok Voice Agent API**, enabling native, real-time bidirectional audio streaming for developers.
99+
- Zhipu AI releases **GLM-4.7**, an open-weights model that reaches the top of global coding and reasoning leaderboards.
100+
- Alibaba open-sources **Z-Image-Turbo**, a highly efficient 6B model, and releases **Qwen-Image-2512**, which specializes in high-fidelity typography and complex visual compositions.
101+
- MiniMax releases **MiniMax-M2.1**, a 200k-context MoE model that rises to the top of web development and coding leaderboards, establishing itself as a leading open model for developers.
102+
- A specialized system by **Poetiq**, powered by GPT-5.2, reportedly solves the **ARC-2** benchmark, marking a major breakthrough in abstract reasoning. (*special*)
103+
104+
---
105+
106+
## דצמבר 2025
107+
- Mistral AI משיקה את משפחת **Mistral 3** (כולל Large ו-Ministral), את **Mistral OCR 3** ואת סדרת **Devstral 2**, מהלך המחזק את מעמדה בתחום המודלים הפתוחים עם יכולות סוכן (Agentic) מתקדמות וכלי פיתוח חדשים.
108+
- OpenAI משחררת את **GPT-5.2**, הכולל את **Codex** לפיתוח תוכנה אוטונומי, ואת **GPT-Image 1.5** התופס את המקום הראשון במדדי הראייה (Vision) מול מודלים מובילים.
109+
- Google מציגה את **Gemini 3.0 Flash** המציב רף חדש ביחס עלות-ביצועים, את סוכן ה-**Deep Research** למחקר עומק וסינתזת מידע, ואת **Gemini 2.5 Flash Audio** לאינטראקציה קולית מהירה.
110+
- Amazon חושפת את סדרת **Nova 2**, ובראשה **Nova 2 Sonic**, מודל Native Speech-to-Speech המאפשר שיחות טבעיות ומלאות הבעה בשיהוי אפסי.
111+
- Runway משיקה את **Gen-4.5**, מודל וידאו חדש המטפס לראש הטבלאות בזכות עקביות תנועה ודיוק גבוה בהבנת הפרומפט.
112+
- xAI משחררת את **Grok Voice Agent API**, המאפשר למפתחים להטמיע שיחות קוליות דו-כיווניות בזמן אמת (Real-time).
113+
- Zhipu AI משחררת את **GLM-4.7**, מודל במשקלים פתוחים המגיע למקום הראשון בלוחות הדירוג העולמיים למשימות קוד והיגיון.
114+
- Alibaba משיקה את **Z-Image-Turbo**, מודל 6B מהיר ויעיל בקוד פתוח, ואת **Qwen-Image-2512** המציג יכולות מתקדמות בטיפוגרפיה וקומפוזיציה ויזואלית מורכבת.
115+
- MiniMax משחררת את **MiniMax-M2.1**, מודל MoE עם חלון הקשר של 200k, המטפס לראש הטבלאות במשימות פיתוח ו-Web וממצב את עצמו ככלי מוביל למפתחים.
116+
- מערכת ייעודית של **Poetiq**, המבוססת על GPT-5.2, מדווחת על פתרון מלא של מדד ה-**ARC-2**, הישג חסר תקדים בהסקה מופשטת. (*special*)

0 commit comments

Comments
 (0)