Create articles from any YouTube video or use our API to get YouTube transcriptions
Start for freeOpenAI's Latest Model Releases
OpenAI has been busy this week with several new model announcements and updates:
Retiring GPT-4 and GPT-4.5
OpenAI announced they will be retiring the GPT-4 model that was released in spring 2023 on April 30th. They are also phasing out the GPT-4.5 model, which was only released in late February 2025.
Introducing GPT-4.1 Series
To replace these models, OpenAI has rolled out GPT-4.1, GPT-4.1 Mini, and GPT-4.1 Nano. These new models are currently only available via API, not in the ChatGPT interface. Some key points about GPT-4.1:
- Faster response generation compared to previous "thinking" models
- Improved performance on coding tasks
- Massive 1 million token context window
- Significantly lower pricing compared to GPT-4.5
New "Thinking" Models: GPT-3 and GPT-4 Mini
OpenAI also released GPT-3 and GPT-4 Mini, which are available within ChatGPT. These are "thinking" models that reason through prompts before responding. Key capabilities include:
- Integration of images into reasoning process
- Ability to transform images (rotate, zoom, etc.)
- Full access to ChatGPT tools and custom tools via API
- Web search functionality during reasoning
Impressive Math Performance
When combined with Python tools, these new models show remarkable math abilities:
- GPT-3 with Python: 95.2% on competition math
- GPT-4 Mini with Python: 98.7% on competition math
- GPT-3 with Python: 98.4% on 2025 competition math
- GPT-4 Mini with Python: 99.5% on 2025 competition math
Novel Idea Generation
According to reports, these new models may be capable of generating novel ideas in fields like material science and drug discovery. This represents a significant step towards artificial general intelligence (AGI).
Image Location Identification
The new models have shown an uncanny ability to identify locations from images with high accuracy, even pinpointing latitude and longitude coordinates.
Codeex CLI and Potential Acquisition
OpenAI released Codeex CLI, an open-source coding tool for use in terminals. There are also rumors that OpenAI is in talks to acquire Windsurf, an AI coding agent interface, for around $3 billion.
Upcoming Releases
Sam Altman has hinted at the release of GPT-3 Pro for ChatGPT Pro subscribers in the coming weeks. There are also rumors of OpenAI developing an X-like social media network.
Image Generation Updates
ChatGPT now has a new library feature allowing users to view and scroll through their AI-generated images.
Google's Latest AI Developments
Gemini 2.5 Flash
Google has introduced Gemini 2.5 Flash, a new large language model with some notable features:
- First fully hybrid reasoning model (thinking can be turned on/off)
- Lower cost compared to competitors like GPT-4 Mini and Claude 3.7
- Competitive performance in science, mathematics, and coding
- Available for testing on Google's AI Studio (ai.dev)
Dolphin Gemma
Google unveiled Dolphin Gemma, an AI model designed to study dolphin communication. This open-source model aims to push the boundaries of interspecies communication.
Video Generation in Gemini
Google has integrated its V2 video generation capabilities into Gemini and Whisk platforms. Users can now generate videos directly from text prompts within these interfaces.
Free Access for College Students
Google is offering US college students free access to Gemini Advanced, Notebook LM Plus, and 2TB of storage for the current and next school year.
Anthropic's Claude Updates
Claude has received some new features:
- Research functionality that integrates with Google Workspace
- Google Workspace integration for email, calendar, and documents
- Upcoming voice mode feature
Grock from XAI
Grock has introduced new features:
- Grock Studio: Adds code execution and Google Drive support
- Memory feature: Remembers past conversations for context
AI Video Generation Breakthroughs
Clling 2.0
Clling has launched version 2.0 of their AI video generation model with significant improvements:
- Better adherence to actions and camera movements
- Enhanced dynamics and more natural range of motion
- Improved visual style consistency and dramatic expressions
- Ability to swap actors in existing film scenes
Arcads.ai
Arcads.ai introduced a new model with gesture control, allowing users to prompt AI actors to generate specific expressions and actions.
Luma Dream Machine
Luma Dream Machine now offers the ability to adjust camera angles in generated videos, providing a wide range of shot types and movements.
Other AI News
- Crisp introduced an accent conversion feature for call centers
- Netflix is testing an OpenAI-powered search engine for personalized recommendations
- Apple and Meta are competing to release AR glasses
- Google demonstrated impressive AR glasses with translation and navigation capabilities
The AI landscape continues to evolve rapidly, with new models and features being released at an unprecedented pace. As these technologies become more sophisticated and accessible, we can expect to see even more groundbreaking applications and use cases emerge in the near future.
To stay up-to-date with the latest AI news and tools, be sure to check out futuretools.io and subscribe to their free AI newsletter for twice-weekly updates on the most important developments in the field.
Article created from: https://youtu.be/ioyAYk5G68Q?feature=shared
