AI posts
AI release notes
Official updates from Anthropic, OpenAI, Google, xAI, Mistral, DeepSeek and Meta, newest first. Open any update for a short AI summary and the full notes.
Compiled by the Toolstash team · Updated
Anthropic · Claude API
Our prompt generator tool is now available in the Developer Console
Prompt Generator makes it easy to guide Claude to generate high-quality prompts tailored to your specific tasks. Read more in our blog post.
Google · Gemini API
Released Gemini 1.5 Flash in preview
Released Gemini 1.5 Flash (gemini-1.5-flash-latest) in preview.
Google · Gemini app
The Gemini mobile app is available in more languages, plus 1 more change
What: The Gemini app is available in languages like Japanese, Korean, Spanish, Portuguese, and more. Chat to get help with writing, planning, learning and more in a variety of supported languages. Learn more about language, country, and device availability in the Help Center. Why: We believe AI should be easier to access in your language, on the device you use the most.
Mistral AI · Mistral APIAnnouncement
For function calling, tool_call_id must not be null for open-mixtral-8x22b
We released three versions of tokenizers for commercial and open-weight models: check the related guide and repo for more details.
Mistral AI · New modelAnnouncement
New model available: open-mixtral-8x22b
Check the release blog for details.
Google · Gemini API
Released Gemini 1.5 Pro in preview
Released Gemini 1.5 Pro (gemini-1.5-pro-latest) in preview. Released a new text embedding model, text-embeddings-004, which supports elastic embedding sizes under 768. Released the File API for temporarily storing media files for use in prompting. Added support for prompting with text, image, and audio data, also known as multimodal prompting. To learn more, see Prompting with media. Released System instructions in beta. Added Function calling mode, which defines the execution behavior for function calling. Added support for the response_mime_type configuration option, which lets you request responses in JSON format.
Mistral AI · Mistral API
JSON mode now available for all models on La Plateforme
JSON mode now available for all models on La Plateforme.
Google · Gemini API
Added support for tuning Gemini 1.0 Pro in Google AI Studio or with the Gemini API
Added support for tuning Gemini 1.0 Pro in Google AI Studio or with the Gemini API.
Google · Gemini app
A better way to tune the Gemini web app’s responses
What: We’re launching a more precise way for you to tune Gemini’s responses. Starting in English in the Gemini web app, just select the portion of text you want to change, give Gemini some instruction, and get an output that’s closer to what you are looking for. Why: We want to give you more control over your creative process by letting you iterate on content and ideas in the context of the original response.
Mistral AI · Mistral API
API endpoints: We renamed 3 API endpoints and added 2 model endpoints
open-mistral-7b (aka mistral-tiny-2312): renamed from mistral-tiny. open-mixtral-8x7B (aka mistral-small-2312): renamed from mistral-small. New API capabilities: Function calling: available for Mistral Small and Mistral Large. JSON mode: available for Mistral Small and Mistral Large La Plateforme: We added multiple currency support to the payment system, including the option to pay in US dollars. We introduced enterprise platform features including admin management, which allows users to manage individuals from your organization. Le Chat: We introduced the brand new chat interface Le Chat to easily interact with Mistral models. You can currently interact with three models: Mistral Large, Mistral Next, and Mistral Small.
Mistral AI · New model
mistral-small-latest: new model
mistral-medium-latest (aka mistral-medium-2312): old model. mistral-large-latest (aka mistral-large-2402): our new flagship model with leading performance.
Google · Gemini app
Enterprise-grade data protections for Workspace customers is now available in Gemini
What: Available today, the new Gemini Business and Gemini Enterprise plans give Google Workspace customers access to one of Google’s most capable AI models, 1.0 Ultra in Gemini and enterprise-grade data protections. Administrators can manage Gemini settings via the Google Workspace admin console. The Gemini Business and Gemini Enterprise plans are available and optimized in English in over 150 countries and territories. Why: Businesses should be able to use Gemini confidently at work. With the Gemini Business and Gemini Enterprise plans, you can elevate your creativity and productivity with access to one of our most powerful models, 1.0 Ultra, while also trusting that your conversations aren't used to train Gemini models. This is just the beginning with even more innovation to come, upgrade today to Gemini Business or Gemini Enterprise.
Google · Gemini app
Exclusive to Gemini Advanced: Edit and run Python code
What: Exclusive to Gemini Advanced, you can now edit and run Python code snippets directly in Gemini's user interface. This allows you to experiment with code, see how changes affect the output, and verify that the code works as intended. Why: These coding capabilities are particularly beneficial for two main use cases: learning and verification. For example, students can play with Gemini's code examples to better understand how modifications impact the outputs. This interactive learning experience can help you grasp coding concepts more effectively. Or, developers can quickly check if the code generated by Gemini runs correctly before copying it. This saves you time and ensures that the code you use is functional.
Google · Gemini app
Bard is now Gemini, plus 3 more changes
What: Gemini is the best way to get direct access to Google AI. All the collaborative capabilities you know and love are still here, and will keep getting better in the Gemini era. We’ve also evolved the UI to reduce visual distractions, improve legibility, and simplify the navigation. Why: We’re committed to giving everyone direct access to Google AI and, as of this week, every Gemini user across our supported countries and languages has access to Google’s best family of AI models. To better reflect this commitment, we’ve renamed Bard to Gemini.
Google · Gemini app
Create images with Bard, plus 2 more changes
What: You can create captivating images for work, play, and anything in between with Bard. It’s easy to use - simply enter a few words to bring your imagination to life, starting with English prompts. Click ‘Generate more’ for more options and download the ones you like. Learn more. Why: Sometimes you might have a picture in mind and now you can easily bring that to life. Create totally unique images in seconds without having to learn complex software.
Mistral AI · Mistral API
We added token usage information in streaming requests
You can find it in the last chunk returned.
Mistral AI · Mistral API
We have enhanced the API's strictness
Previously the API would silently ignores unsupported parameters in the requests, but it now strictly enforces the validity of all parameters. If you have unsupported parameters in your request, you will see the error message "Extra inputs are not permitted". A previous version of the guardrailing documentation incorrectly referred to the API parameter as safe_mode instead of safe_prompt.
Google · Gemini app
Expanding Bard Extensions to new languages, plus 1 more change
What: You can now use Bard to access helpful information from Google apps and services in Japanese and Korean, as well as English. With Bard Extensions, you can retrieve real-time info from YouTube, Hotels, Flights, and Maps and even allow Bard to retrieve information from your Gmail, Docs, and Drive, so you can easily find, summarize and get answers from your personal content. You're in control of your privacy settings when deciding how you want to use extensions. Why: We're committed to building Bard responsibly, gradually bringing key features to more languages over time.
Google · Gemini API
gemini-pro: New text model for a wide variety of tasks. Balances capability and efficiency
gemini-pro: New text model for a wide variety of tasks. Balances capability and efficiency. gemini-pro-vision: New multimodal model for a wide variety of tasks. Balances capability and efficiency. embedding-001: New embeddings model. aqa: A new specially tuned model that is trained to answer questions using text passages for grounding generated answers. v1: The stable API channel. v1beta: Beta channel. This channel has features that may be under development. GenerateContent is a single unified endpoint for chat and text. Streaming available through the StreamGenerateContent method. Multimodal capability: Image is a new supported modality New beta features: Function Calling Attributed Question Answering (AQA) Updated candidate count: Gemini models only return 1 candidate. Different Safety Settings and SafetyRating categories. Tuning models is not yet supported for Gemini models (Work in progress). Terms Privacy Manage cookies English Deutsch Español – América Latina Français Indonesia Italiano Polski Português – Brasil Shqip Tiếng Việt Türkçe Русский עברית العربيّة فارسی हिंदी বাংলা ภาษาไทย 中文 – 简体 中文 – 繁體 日本語 한국어
Google · Gemini appAnnouncement
Bard is getting its biggest upgrade yet with Gemini Pro
What: Starting today, we’re introducing Gemini Pro in Bard, for Bard’s biggest upgrade yet. We’ve specifically tuned Gemini Pro in Bard to be far more capable at things like understanding and summarizing, reasoning, coding, and planning. You can try out Bard with Gemini Pro for text-based prompts, with support for other modalities coming soon. It will be available in English in more than 170 countries and territories to start, and come to more languages and places, like Europe, in the near future. Why: Today, Google introduced Gemini, the most capable AI model in the world. Gemini unlocks new ways to create, interact and collaborate with Bard.
Google · Gemini app
Expanding Bard’s understanding of YouTube videos
What: We're taking the first steps in Bard's ability to understand YouTube videos. For example, if you’re looking for videos on how to make olive oil cake, you can now also ask how many eggs the recipe in the first video requires. Why: We’ve heard you want deeper engagement with YouTube videos. So we’re expanding the YouTube Extension to understand some video content so you can have a richer conversation with Bard about it.
Google · Gemini app
Bard is available for more age groups, plus 2 more changes
What: Bard is expanding access to teens in most countries around the world, starting with English. We’ve added age-appropriate protections, updated onboarding for teens, and developed experiences geared to empower exploration and learning with Bard. Want to learn more about generative AI and its abilities and limitations? Learn about generative AI. Why: We believe Bard can be a helpful tool for teens when they need a little extra inspiration and motivation on their ideas, hobbies, and plans, or when they want to better understand topics quickly in a style that works for them. Whether it's learning homework concepts or getting support through big milestones like applying to their first job or preparing for college, Bard can help.
Google · Gemini app
Bard’s responses can now appear in real time
What: We’re launching a new setting that lets Bard’s responses show while in progress, so you don’t have to wait for the full response to appear. Why: Bard is here to accelerate your creative process. Now you can start reading responses while they’re still generating, so you can stay in the creative flow and iterate on your ideas faster.
Google · Gemini app
Bard can now summarize more of your emails, plus 1 more change
What: When you use the Workspace Extension, Bard can now summarize more emails at a time. It can also better understand when you ask for recent emails. Why: We are taking a small step to improve quality and make the Workspace Extension more useful by summarizing more of your emails. Sometimes you just need an update, so we’re improving the way Bard can understand when you need the latest emails.
Google · Gemini app
Help Bard improve side by side
What: We’ve added a new way for you to give feedback. When Bard occasionally responds with 2 drafts side by side, select the draft you prefer. You can also indicate no preference or opt out entirely. Why: Real-world feedback helps Bard improve the quality of its responses. Your feedback on Bard drafts contributes to making Bard better for yourself and others.
Google · Gemini app
Introducing Bard’s most capable model yet, plus 5 more changes
What: Your feedback has accelerated Bard’s ability to be more intuitive, imaginative, and responsive than ever before. Whether you want to collaborate on something playful and creative, start in one language and continue in up to 40+ other languages, ask for in-depth coding assistance, or learn about new topics from different points of view, Bard can help you explore possibilities with greater quality and accuracy. Why: We believe in rapid iteration and bringing the best of Bard to the world. Your feedback has helped unlock the most capable version of Bard yet for stronger collaboration in languages and countries all over the world.
Google · Gemini app
Bard is available in new places and languages, plus 6 more changes
What: Bard is now available in over 40 new languages including Arabic, Chinese (Simplified/Traditional), German, Hindi, Spanish, and more. We have also expanded access to more places, including all 27 countries in the European Union (EU) and Brazil. Why: Bard is global and is intended to help you explore possibilities. Our English, Japanese, and Korean support helped us learn how to launch languages responsibly, enabling us to now support the majority of language coverage on the internet.
Google · Gemini app
Bard codes for you for improved math and data analysis, plus 1 more change
What: Starting with English, we’ve updated Bard to detect computational prompts and run code in the background, making Bard better at mathematical tasks, coding questions, and string manipulation. Why: We’re supercharging your ability to analyze data by making Bard better at computation, like finding how many times a certain number shows up in a data set. Bard won’t always get it right, but will keep improving with your feedback.
Google · Gemini app
More relevant responses with location info
What: Bard can start providing more relevant responses if you choose to let it use your device’s precise location. You can manage your preferences in location settings. Why: Precise location helps Bard provide more relevant responses about restaurants near you and many other things about your area.
Google · Gemini app
Bringing images to Bard
What: Starting with English responses, Bard can now bring in images from Google Search, so you can get helpful responses with visuals. You can also ask Bard for images directly. Bard will show a source for each image. Why: Images can help you communicate your ideas more effectively. They can bring concepts to life, make recommendations more persuasive, and enhance responses when you ask for visual information.