The AI model is cranked out at an eye-opening pace, from big companies like Google to startups like Openai and humanity. Tracking the latest is overwhelming.
In addition to the confusion, AI models are often promoted based on industry benchmarks. However, these technical indicators reveal little about how people and businesses actually use them.
To get through the noise, TechCrunch has compiled an overview of cutting-edge AI models released since 2024. We’ve recorded details about how they’re used and what they’re best suited to. This list will also be updated with the latest releases.
There are literally more than a million AI models. For example, hugging your face Hosts over 1.4 million people. Therefore, this list may miss models that somehow improve performance.
AI model released in 2025
Google Gemini 2.5
Gemini 2.5 Pro Experimental, the inference model, according to Google, is excellent at building web apps and code agents. However, compared to Claude Sonnet 3.7, performance is degraded with one common coding benchmark. This model requires a $20 monthly Gemini Advanced subscription.
chatgpt-4o image generator
Openai upgraded the existing GPT-4O model to generate images as well as text. The souped model quickly went viral, despite obvious copyright concerns, to convert the image into Studio Ghibli-style anime. To access the GPT-4o, you will need at least $20 per month ChatGPT and subscription.
Stable AI Stable Virtual Camera
Image Generation Launch Stability AI launched a model the company says can generate 3D scenes and camera angles from a single 2D image. However, it still struggles with scenes that are characterized by more complex elements like humans and moving water. This model can be used for non-commercial research use on Huggingface.
Cohere’s Aya Vision
Cohere has released a multimodal model called AYA Vision, claiming it is the perfect fit for the class, including captioning images and answering photo questions. Unlike other models, it also excels in languages other than English. Available Free on WhatsApp.
Openai’s GPT 4.5 “Orion”
Openai calls Orion the biggest model ever, promoting its powerful “world knowledge” and “emotional intelligence.” However, performance is degraded on certain benchmarks compared to new inference models. Orion is available to Openai subscribers for $200 per month.
Claude Sonnet 3.7
Humanity says this is the industry’s first “hybrid” reasoning model. Because it fires a quick answer and allows you to really think about things when you need them. You can also control the user how long a model can think of. All humanity. Sonnet 3.7 is available to all Claude users, but heavier users will need a Pro Plan that costs $20 per month.
Xai’s Grok 3
The Grok 3 is the latest flagship model of startup Xai, discovered by Elon Musk. the It was claimed It outperforms other major models of mathematics, science and coding. The model requires X Premium ($50 a month). Found Grok 2 Leaned Leaded, Musk Pledge It is not yet clear whether it has been achieved to turn Glock into a more “political neutral.”
Openai O3-Mini
This is OpenAI’s latest inference model, optimized for STEM related tasks such as coding, mathematics, and science. the Not the most powerful of Openai It’s a model, but it’s small so it’s a company. say That’s a significantly lower cost. It is free to use, but requires a heavy user subscription.
Openai Deep Research
Openai’s deep research is designed to provide in-depth research on topics with clear citations. This service is available only with ChatGPT’s $200/month Pro subscription. Openai I recommend it For everything from science to shopping research, however, it should be noted that hallucinations remain an AI problem.
Mistral Le Chat
Mistral has launched the APP version of Le Chat, a multimodal AI personal assistant. Mistral claim LE Chat responds faster than any other chatbot. There is also a paid version Latest Journalism From AFP. Test from Le Monde I found Le Chat’s performance to be impressive, but it got more errors than ChatGpt.
Openai Operator
The Openai operator is intended to be a personal intern who can do things independently to help you buy groceries. You will need a $200 ChatGPT Pro subscription per month. AI agents have many promises, but they are still experimental: Washington Post Reviewer The operator says We have independently decided to order a dozen eggs for $31 paid by Reviewers credit card.
Google Gemini 2.0 Pro Experimental
Google Gemini’s long-awaited flagship model states it is excellent at coding and understanding general knowledge. It also has a very long context window of 2 million tokens, which is useful for users who need to quickly process large chunks of text. This service requires (at least) $19.99 a month on a Google One AI Premium subscription.
AI model released in 2024
Deepseek R1
This Chinese AI model took Silicon Valley by storm. Deepseek’s R1 works well in coding and mathematics, but the open source nature means that anyone can run locally. Plus, it’s free. However, R1 faces an upward ban as it consolidates Chinese government censorship and could send user data back to China.
Gemini Deep Research
Deep Research brings together Google search results into simple, well-cited documents. service It is useful for students and others who need a brief research summary. However, its quality is not as good as actual peer-reviewed papers. Deep Research requires a $19.99 Google One AI Premium subscription.
Metalama 3.3 70b
This is the latest and most advanced version of Meta’s open source Llama AI model. Meta is advertised This version It is the cheapest and most efficient, especially for mathematics, general knowledge, and the following instructions. It’s free and open source.
Openai Sora
SORA is a model that creates realistic videos based on text. You can generate the entire scene rather than just a clip, Openai admits Often generating “unrealistic physics.” Currently, it is only available in paid versions of ChatGpt starting with Plus. This is $20 a month.
Alibaba QWen QWQ-32B-PREVIEW
This model is one of the few models comparable to Openai’s O1 in benchmarks in a particular industry, and is excellent at mathematics and coding. Ironically, with regard to the “inference model,” there is “a room for improvement in common sense inference.” Alibaba says. It also includes a test show from the Chinese government’s censorship, TechCrunch. It’s free and open source.
Humanity’s computer use
Claude’s computer usage is intended to control the computer to complete tasks such as coding and booking plane tickets, and to become the predecessor of Openai’s operators. However, using a computer It remains in beta. The price is via the API, $0.80 per 100 tokens for input and $4 per 1 million tokens for output.
Xai’s Grok 2
Elon Musk’s AI Company, Xai launches an expanded version of flagship Grok 2 Chatbot It claim It’s 3x faster. Free users are limited to 10 questions every two hours on Grok, while X’s Premium and Premium+ plan subscribers enjoy higher usage restrictions. Xai also launched Aurora, an image generator that generates highly light-electoral images with graphics and violent content.
Openai O1
Openai’s O1 family aims to produce better answers by “thinking” through responses through hidden inference functions. The models are excellent in coding, mathematics and safety. Openai’s claimHowever, there is also a problem with trying to deceive humans. To use O1, you must subscribe to chatgpt Plus. This is $20 a month.
Claude Sonnet of Mankind 3.5
Claude Sonnet 3.5 is the best model human claim in its class. It has become known for its coding features and is considered a technology insider Selected chatbot. The model is free to access on Claude, but heavy users will need a $20 monthly Pro subscription. I understand the image, but I can’t generate it.
Openai gpt 4o-mini
Openai is promoting the GPT 4o-Mini as the most affordable and fastest model ever thanks to its small size. That is meant To enable a wide range of tasks, such as turning on a customer service chatbot. This model is available in the free tier of ChatGpt. It is suitable for a large number of simple tasks compared to more complex tasks.
Cohere Command R+
Cohere’s Command R+ Model Excellent for complex, searched, generated (or RAG) applications in the enterprise. This means you can find and quote certain information very well. (Rag’s inventor It actually works with Cohere. ) Still, RAG does not completely solve the problem of AI hallucination.
Arabic
Dutch
English
French
German
Italian
Portuguese
Russian
Spanish