Google Cloud made a series of AI announcements today, with new models available in Vertex AI, updates to the Gemini API, and new languages in Google Translate enabled by AI. Developers can now take advantage of the 2 million token context window on Gemini 1.5 Pro without needing to be patient on a waitlist. Plus, you can now apply to be one of a limited number of users of Google's newest image generator, Image 3, which can create photorealistic images for marketing or corporate presentations.
Google's newest AI is open to the public
Currently, there are new or higher-performing models of several Google AIs available on the Vertex AI platform:
- Gemini 1.5 Flash, a relatively compact model with a 1 million token context window, is generally available.
- Gemini 1.5 Pro is now generally available.
- Image 3 is in preview. Apply here.
“Gemini 1.5 Flash makes it easier for us to continue our scaling phase of applying generative AI to high-volume tasks without sacrificing output quality or context window, even for multimodal use cases,” said JC Escalante, global head of generative AI at market research firm Ipsos, in a Google press release.
Vertex AI now offers or will soon offer:
- The light variant of Gemini Gemma 2:Generally available from Vertex AI next month in two sizes: 9 billion parameters and 27 billion parameters
- Anthropic's Claude Sonnet 3.5, out now.
- Context caching, a technique used to create greater speed and lower cost for AI requests that use repetitive content, is now in public preview for Gemini 1.5 Pro and Flash.
- Provisioned Throughput, a Vertex AI feature for workloads provisioned on Gemini models, is now generally available to allowlisted users.
- Grounding is now available for greater accuracy, where AI can compare your information with that of Google Search. The Grounding feature from third parties, such as Thomson Reuters, is expected to roll out starting next quarter.
- High-fidelity grounding mode, which combines Gemini 1.5 Flash with enterprise data, is now in experimental preview.
Vertex AI is available in a wide variety of geographic regions.
SEE: Here are five ways to search the web with generative AI.
Gemini API can now execute code execution and more
Code execution is now possible in Gemini 1.5 Pro and Gemini 1.5 Flash, allowing developers to run Python within the model and experiment by allowing generative AI to iterate and learn from the code. It can be accessed through the Gemini API or Google AI Studio.
Additionally, Gemini API users can now:
- Use the full 2 million token window on Gemini 1.5 Pro.
- Use context caching for both Gemini 1.5 Pro and 1.5 Flash.
- Experiment with Gemma 2 in Google AI Studio.
Cantonese and 109 other languages have been added to Google Translate
Google has used the PaLM 2 language model to add 110 languages to the public Google Translate service; this is the largest expansion ever of this service. One highlight is Cantonese, a language that Google has struggled to find data for to add to Translate in the past because it “often overlaps with Mandarin in writing.”
PaLM 2 has allowed Google to more efficiently add more languages similar to each other, said Google senior software engineer Isaac Caswell in a press release about this expansion of Google Translate.
