Tag: artificial intelligence

Google has moved Gemini 4 Argon from a limited early-access model to broad commercial availability. Since October 13, 2026, developers and enterprises can use it through the paid Gemini API, while Google AI Pro and Ultra subscribers can select it in the Gemini app and AI Mode in Search. Ultra subscribers also receive access through Google Antigravity.

The launch is aimed at difficult work rather than quick everyday questions. Google positions Argon for long-running software engineering, multimodal document analysis, science, finance, legal research and defensive cybersecurity. Its headline specification is a unified input-and-output limit of 1,048,576 tokens, but access, pricing and feature support differ sharply between the consumer app and the API.

Gemini 4 Argon at a glance

• Availability: globally available from October 13, 2026, through the paid Gemini API and selected Google AI subscription products.
• Consumer access: Google AI Pro and Ultra subscribers can use Argon in the Gemini app and AI Mode in Search.
• Developer access: the stable API model ID is gemini-4-argon in Google AI Studio and the Gemini API paid tier.
• Context and output: a unified 1,048,576-token limit, with synchronous output capped at 65,536 tokens unless long decode is used.
• Introductory API price: $2 per million input tokens and $10 per million output tokens through November 30, 2026.
• Standard price beginning December 1: $4 per million input tokens and $20 per million output tokens.
• Supported inputs: text, images, video, audio and PDF; output is text.
• Important omissions: no image generation, audio generation or Live API support on this model.

Who can use Gemini 4 Argon now

The simplest route is the Gemini app. Google AI Pro and Ultra subscribers can choose Argon for complex prompts without configuring an API. The same subscriber tiers can use it in AI Mode in Search, where web grounding and follow-up exploration are part of the interface. Free Gemini accounts are not listed for Argon access in Google’s launch announcement.

Developers can call the stable gemini-4-argon model through the paid Gemini API in Google AI Studio. Google also lists enterprise access through its business products and Gemini platform. Antigravity, Google’s agentic coding environment, is narrower: the announcement identifies Google AI Ultra subscribers for consumer access there, while enterprises have a separate path.

Availability does not mean unlimited use. Consumer plans can apply usage caps, and API customers pay for tokens and any other billable services. Teams should confirm the plan, region and account policy shown in their own Google console before committing a production workflow.

What Gemini 4 Argon is built to do

Argon is Google’s flagship reasoning model for tasks that require many connected steps and substantial context. Google highlights large-codebase refactoring, professional document analysis, scientific reasoning, autonomous agents and defensive security work. The model can accept text, images, video, audio and PDFs, which makes it useful when a project mixes source code, specifications, diagrams, recordings and long reference documents.

Its tool support includes code execution, function calling, file search, URL context, structured outputs, Search grounding and Google Maps grounding. Computer use is available in preview. Thinking can be set to low, medium or high, giving developers a way to trade speed and cost against deeper reasoning.

Argon is not a universal media model. It returns text and does not support image generation, audio generation or the Live API. Teams that need real-time voice interaction or generated media will still need another Gemini model or a multi-model workflow.

The one-million-token limit needs careful reading

Google describes Argon as having an industry-leading one-million-token unified input-output limit. Unified means the input and generated response share the same overall budget. A very large prompt therefore leaves less space for the answer, and a long answer leaves less room for source material.

The API reference lists a default synchronous maximum output of 65,536 tokens. Reaching output lengths closer to the full one-million-token ceiling requires Google’s long-decode continuation workflow. Developers should not assume that a normal single request will return a book-length response merely because the headline limit is one million.

A large context window also does not guarantee perfect recall. Important instructions can be diluted by redundant files, conflicting policies or repeated drafts. Strong workflows still need a clear task, organized source material, explicit output requirements and validation against the original evidence.

Gemini 4 Argon pricing before and after November 30

The launch price is temporary. Through November 30, 2026, Google charges $2 per million input tokens and $10 per million output tokens for the public Argon model. Cached input receives a 95 percent discount, although the model page says Argon supports implicit caching rather than explicit caching.

Beginning December 1, the published rate doubles to $4 per million input tokens and $20 per million output tokens. That date matters for pilots: a workflow that looks affordable in October or November can cost twice as much at the same token volume in December.

For a simple illustration, one million input tokens plus 100,000 output tokens would cost about $3 during the introductory period: $2 for input and $1 for output. At the standard rate, the same uncached request would cost about $6. This example excludes other platform charges and assumes the published public-model rates.

• Measure real prompt and output sizes before estimating a monthly budget.
• Remove duplicate files and stale context instead of paying to resend them.
• Test whether a lower thinking level is sufficient for routine work.
• Track the December 1 price change separately from usage growth.
• Use production limits, logging and human approval for costly agent loops.

Gemini app, AI Mode, API and Antigravity are different products

In the Gemini app, Argon is packaged as a model choice for subscribers. Users provide prompts and files through Google’s interface, while plan limits and product settings control availability. AI Mode in Search adds a research-oriented surface that can ground responses in current web results.

The API gives developers much more control over tools, thinking level, structured outputs and application logic, but it also creates responsibility for authentication, budgets, logging, safety and data handling. Antigravity is designed for agentic coding and longer software tasks. It is not simply another name for the Gemini chat interface.

A team should choose the surface around the job. Occasional analysis may fit the Gemini app. A repeatable business workflow belongs in the API only after the team has tested costs, permissions and failure handling. Large autonomous coding tasks may justify Antigravity, but they still require review before code is merged or deployed.

Capabilities developers should verify before migrating

• Multimodal inputs are supported, but generated output is text only.
• Code execution, file search, function calling and structured outputs are supported.
• Search and Google Maps grounding are supported; URL context is also available.
• Computer use is marked preview and should not be treated as a fully stable production dependency.
• Implicit caching is supported, while explicit caching is not listed for Argon.
• Batch, Flex and Priority inference are listed as consumption options.
• Live API, image generation and audio generation are not supported.

These differences can break an application that swaps only the model ID. A system designed for explicit caches, live audio or generated images needs architectural changes rather than a one-line migration. Google’s latest-model guide is the correct place to check request changes and continuation behavior before production use.

How to interpret Google’s performance claims

Google reports that Argon ranks first on the Vals Index and cites internal or partner results across coding, science and cybersecurity. Examples include faster codebase refactoring, improved framework performance and assistance with vulnerability discovery. These are useful signals, but they are not a promise that every customer will reproduce the same gains.

Workload design matters. A model can perform strongly on a curated benchmark while struggling with an organization’s private terminology, incomplete documentation or unusual approval process. The safest evaluation uses representative tasks, blinded scoring where practical and a fixed comparison against the current system.

Google also reports improvements from internal Antigravity use, including more agentic code submissions and fewer rollbacks. Because those figures come from Google’s own environment, buyers should treat them as launch evidence rather than an independent service-level guarantee. Reuters separately reported the wider October 13 commercial launch and its focus on developers and businesses.

Safety and cybersecurity limits still matter

Argon’s ability to inspect code and work across tools raises the consequences of a bad instruction. Google says the model received new jailbreak defenses, prompt-injection training and agent monitoring. The company reports a 0.7 percent attack success rate on a multi-attempt indirect prompt-injection benchmark and says shortcut attempts fell to 2 percent in selected coding evaluations.

Those are vendor-reported evaluation results, not proof that prompt injection is solved. A production agent can still encounter malicious instructions in webpages, email, documents or source repositories. Tool permissions, network access and approval gates should be limited to the smallest scope the workflow needs.

For consequential changes, keep a human checkpoint before deployment, deletion, permission changes, purchases or messages to third parties. Security teams should also separate defensive testing from unrestricted exploit generation and follow the access conditions of Google’s Fairwind Program.

Should existing Gemini users switch

Subscribers should choose Argon when the task is genuinely difficult: a large repository, a long multimodal evidence set, a multi-stage research problem or an analysis that benefits from extended reasoning. It is likely excessive for rewriting a short email, summarizing one page or answering a simple factual question.

API users should run a controlled trial before migrating traffic. Compare quality, latency, token use, tool reliability and total cost with the current model. Recalculate the budget at the December rates, not only the introductory price. If a smaller model meets the quality target, routing easy requests away from Argon can reduce cost and delay.

Organizations in legal, financial, medical or security-sensitive work should keep domain review. Google describes strong professional capabilities, but the output remains generated text and can contain mistakes, unsupported conclusions or outdated assumptions.

A practical first-week test plan

• Choose five to ten real tasks that are difficult for the current model.
• Define success criteria before testing, including factual accuracy, completeness, latency and cost.
• Run the same inputs through Argon and the current baseline without changing the rubric.
• Inspect citations, calculations, code changes and tool actions against primary evidence.
• Record token use under both introductory and December pricing.
• Keep permissions read-only during early agent tests whenever possible.
• Move only the workloads with a meaningful, repeatable improvement.

A relevant resource for AI-assisted coding

Readers building their first agentic development workflow can look up AI Vibe Coding by Jordan R. Ashford on Amazon. The book’s verified ISBN is 9798253607541; confirm the seller, edition and price before ordering.

As an Amazon Associate, ReaderSpice earns from qualifying purchases.

Bottom line

Gemini 4 Argon is now a real option for paid subscribers, developers and enterprises, with a combination of deep reasoning, multimodal input, broad tool support and an unusually large unified token budget. The most important limits are easy to miss: normal synchronous output is much smaller than the headline ceiling, several media and real-time features are unsupported, and the introductory API price ends after November 30.

Use Argon for work that earns the extra reasoning and context. Test it on representative tasks, validate the output, keep agent permissions narrow and budget at the December rate before making it a default. For simple prompts, a faster and less expensive model may remain the better choice.

Sources

Google Blog: Gemini 4 Argon is now available for developers, enterprises and subscribers

Google AI for Developers: Gemini 4 Argon model specifications

Google AI for Developers: What’s new in Gemini 4 Argon

Google DeepMind: Gemini 4 Argon model card

Google Blog: Gemini 4 Argon availability in the Gemini app

Reuters: Google launches Gemini 4 Argon for developers and businesses

Google has started rolling out Skills in the Gemini app, giving users a way to save reusable instructions instead of rebuilding the same prompt for every chat. The feature can combine focused directions with reference files and can be called manually or applied when Gemini decides a skill matches the task.

The broader workplace launch is not moving on the same schedule. Google’s October 13 update says Skills in Workspace Studio and Workspace apps are delayed until further notice, replacing the earlier plan to finish Rapid Release deployment by October 12 and begin Scheduled Release on October 19. That distinction matters: seeing Skills in the Gemini app does not mean they are ready inside Docs, Gmail or Workspace Studio.

Gemini Skills rollout at a glance

• Skills began a gradual Gemini-app rollout on October 13, 2026, with completion expected by mid-November.
• Google says the Workspace rollout for both Rapid Release and Scheduled Release domains is delayed until further notice.
• A skill stores reusable instructions and can include reference files, reducing repeated setup for recurring tasks.
• Gemini-app Skills and Workspace Skills do not currently sync; users must create them separately on each surface.
• Gems remain available during the transition, but Google plans to replace them with Skills on a staged timeline.
• Sharing and Google Drive integration for Gemini-app Skills are still promised for the coming weeks, not available everywhere today.

What Gemini Skills actually do

A Skill is a reusable set of instructions that tells Gemini how to handle a particular kind of work. Instead of pasting a brand voice, formatting checklist or analysis method into every conversation, a user can save that guidance once and invoke it again. Google’s help pages compare Skills to focused cheat sheets: they should capture the process, preferences, output format and common mistakes for one recurring job.

Skills can be created from a chat, from the Skills page or from a template. They can also include supported reference files such as text documents, PDFs or images. Gemini can apply a relevant Skill automatically, or the user can choose one directly. Multiple Skills can be combined in a single task, which is the biggest functional difference from a standalone Gem.

The underlying format is Markdown-based and uses a SKILL.md file. That open structure is designed to make a Skill portable and inspectable, although Google’s products still control where it can run and which account features are available.

How to create and run a Skill in the Gemini app

• Open Gemini on the web, in the mobile app or in the Mac app, then find Customize in the sidebar and choose Skills.
• Select Create with Gemini, start from a recommended template, use a blank template or upload a folder containing a SKILL.md file and optional references.
• Give the Skill a narrow, action-oriented name and a description that makes its trigger clear.
• Write short instructions covering the steps, required format and errors Gemini should avoid.
• Test the Skill on a real task, inspect the result and revise instructions that are vague or easy to misread.
• Invoke the Skill from the prompt interface or allow Gemini to use it when the task matches.

The rollout is gradual, so the Skills menu may not appear immediately. Google’s Gemini help page also says the feature is currently reaching personal Google Accounts first, even though the Workspace announcement lists eventual Gemini-app availability for all Workspace customers. That documentation mismatch is a strong reason to treat account access as phased rather than guaranteed on a specific day.

Why the Workspace launch is delayed

Google has not identified a bug, security incident or revised launch date. Its official explanation is that more work is needed to launch Skills in Workspace safely and reliably. The status applies to both Rapid Release and Scheduled Release domains, and the company says it will publish an updated timeline when one is confirmed.

That means administrators should discard the old October 12 and October 19 milestones when planning training or migration. A domain that does not show Skills in Workspace is not necessarily misconfigured. The feature may simply not be available yet, even if a personal Gemini account on the same device already has it.

The delay also affects workflows that expected to build or share Skills inside Workspace Studio or Workspace apps. Teams can prepare instructions and test the Gemini-app version, but should avoid promising a production Workspace launch date until Google posts a replacement schedule.

Gemini-app and Workspace Skills are separate

Google says Skills do not currently synchronize between the Gemini app and Workspace apps. A Skill created in Gemini chat will not automatically appear in Docs, Gmail or Workspace Studio. If a team wants the same behavior in both environments, it must recreate the Skill separately after Workspace support becomes available.

The separation has practical consequences. Changes made to one copy will not automatically update the other, reference files may need to be attached again and access controls can differ. Teams should keep a canonical set of instructions in a controlled document or repository so that both versions can be rebuilt consistently.

Sharing is also incomplete in the initial Gemini-app rollout. Google says sharing and Google Drive integration are coming in the weeks after launch. Until those features arrive, an individual Skill should not be mistaken for a centrally managed Workspace asset.

What happens to Gems

Gems are not disappearing immediately. Users can continue creating and using them in the Gemini app during the transition, but new Workspace Studio flows can no longer be created with the “Ask a Gem” step. Existing flows using that step continue to work for now.

Google plans to auto-migrate remaining Gemini-app Gems into draft Skills when Gem support ends. Users who want to move earlier can manually recreate a Gem as a Skill. Because Gemini-app Skills do not sync to Workspace, an early manual migration does not create a Workspace copy.

The transition dates that still matter

• November 17, 2026: Gems move to the Settings panel in the Gemini app, while creation, editing and use continue.
• No sooner than March 1, 2027: Business and enterprise users lose the ability to create, edit or use Gems; remaining Gems are converted to draft Skills.
• No sooner than June 1, 2027: The corresponding Gems shutdown applies to education users, including supported Classroom and Gemini LTI experiences.

Google says administrators will receive at least 30 days’ notice before automatic migration. These are earliest dates rather than promises that every change will occur on the first possible day.

Admin controls, eligibility and age limits

The Gemini-app rollout is designed for personal accounts and Workspace customers, but Google says there are no administrator controls for Skills inside the Gemini app. Workspace Skills will have separate administrative controls once they launch. That difference can matter for organizations that allow Gemini access but need to govern reusable instructions or shared reference material.

Google lists future Workspace Skills availability for Business Starter, Standard and Plus; Enterprise Starter, Standard and Plus; Google AI Pro for Education; and AI Expanded Access. Workspace Skills are limited to users over 18, while the Gemini-app announcement says Skills will be available to users of all ages.

Administrators should review who can access Gemini, which data may be attached as references and how reusable instructions fit retention and confidentiality policies. A Skill can make a workflow more consistent, but it can also repeat a bad rule or expose a reference file to the wrong audience if governance is unclear.

Three practical Skills worth building first

A repeatable document reviewer

Create a Skill that checks a draft for missing evidence, unclear claims, inconsistent terminology and a required output structure. Keep it focused on review rather than asking it to rewrite everything automatically.

A meeting-to-action converter

Use a Skill to turn pasted meeting notes into decisions, owners, deadlines and unresolved questions. Require the output to distinguish explicit commitments from inferred next steps so reviewers can spot uncertainty.

A consistent research brief

Define the sources to prioritize, the date range to check, the fields to capture and the format for separating confirmed facts from assumptions. This is more reliable than repeatedly asking for “a summary” without a standard.

Start with one narrow workflow, test it against several examples and revise the instructions when the output fails. Google’s own guidance recommends short, focused Skills rather than large manuals covering unrelated jobs.

A practical device for Gemini Skills

Skills are rolling out in Gemini on mobile as well as the web, so a current Android phone can be useful for testing reusable instructions away from a desk. Readers comparing compatible hardware can view the unlocked Google Pixel 11 on Amazon.

ReaderSpice may earn a commission from qualifying purchases made through affiliate links, at no additional cost to you.

Bottom line

Gemini Skills are now a real product rollout, but only part of Google’s original schedule is active. Personal Gemini users may begin seeing reusable, stackable instructions through mid-November, while Workspace customers must wait for a new launch timeline.

The safest plan is to experiment without forcing a migration. Keep Gems working where they are still needed, document reusable instructions in a neutral source, and do not assume a Skill in Gemini chat exists in Workspace. When Google resumes the Workspace rollout, teams can recreate tested Skills with clearer governance instead of rebuilding them under deadline pressure.

Sources

Google Workspace Updates: Skills rollout status and Workspace delay

Google Workspace Updates: Introducing Skills and the Gems transition

Google Blog: Let Skills in Gemini tackle repetitive tasks

Gemini Apps Help: Create and manage Skills

Gemini Apps Help: Write effective Skills

Google is bringing two of Gmail’s most ambitious Gemini features to more paid Workspace accounts. Gmail Live turns inbox search into a spoken conversation on Android and iPhone, while AI Inbox builds a separate view of priority topics, deadlines and suggested to-dos drawn from email.

The October 2026 expansion can save time, but availability is unusually specific. Both features are currently limited to the United States and English, require eligible plans and depend on Gmail smart-feature settings. Google also labels both experiences as beta, so readers should verify important answers against the original messages.

Gmail Live and AI Inbox at a glance

• Gmail Live lets users ask spoken questions about their email and hear synthesized answers on supported Android phones and iPhones.
• Each Gmail Live turn can show an email source card, giving users a path back to the messages behind an answer.
• AI Inbox extracts suggested to-dos, deliverables and deadlines, then summarizes priority topics across conversations.
• Both features currently require U.S. availability, English language settings, eligible plans and Gmail smart features.
• Google is using gradual rollouts, so an eligible account may wait as long as 15 days—or longer for some editions—before the features appear.

Gmail Live turns inbox search into a conversation

Gmail Live is designed for questions that would otherwise require several searches or long email threads. A user can ask for a flight gate, an order update, a training deadline or the latest message from a manager, then interrupt, refine the question or follow up without starting a new search.

Google says the feature uses current Gemini models to provide low-latency, inbox-aware voice search. Answers are spoken aloud, making the experience useful when typing is inconvenient. The feature is available inside the Gmail mobile app rather than on the desktop web interface.

This is not a hands-free license to trust every response. Email can contain old reservations, revised deadlines or several messages with similar subjects. Before acting, open the cited message and confirm that the sender, date, attachment status and latest reply match the answer.

Source cards make Gmail Live answers easier to audit

Google now surfaces an email source card prominently at each conversational turn. Users can tap a link inside a response to open the exact email or open the Sources card to see all cited messages. Selecting an email ends the Live session, while a View search page option returns the user to Gmail’s standard results.

That source path is the most important reliability feature in the update. If Gmail Live says a payment is due or a flight leaves from a certain gate, the original message remains the authoritative record. The source card also helps reveal when an answer blended details from more than one thread.

Google’s rollout post says users can choose a preferred Gmail Live voice. At publication time, Google’s Android help page still says the voice cannot be changed. That documentation mismatch likely reflects a staged interface update, so do not assume the voice selector has reached every account yet.

How to start Gmail Live

• Open the Gmail app on a supported Android phone or iPhone and sign in to an eligible account.
• Look for the Gmail Live entry point near the top of the app or in the search area.
• Grant microphone permission the first time Gmail asks, then accept the displayed terms.
• Ask a question aloud, interrupt to refine it or select a suggested prompt.
• Use the Sources card to inspect the messages behind the response before acting.

Google’s help page says Gmail Live does not currently work on Android tablets. Work and school users also depend on their administrator allowing Gemini beta access, while personal accounts need Google AI Plus, Pro or Ultra. Smart features in Workspace and other Google products must be enabled.

AI Inbox builds a prioritized view of work and deadlines

AI Inbox appears as a separate destination in Gmail’s menu. Its Suggested to-dos section extracts high-priority tasks from incoming messages and can emphasize a deadline or required action. Topics to catch up on summarizes important projects, travel plans, reservations, purchases and other threads that may be scattered across the inbox.

Users can open the related message, reply, mark a to-do complete or remove an inaccurate item. Feedback controls allow a positive or negative rating, and Google includes a report-abuse option for unsafe suggestions.

The feature may take time to populate. Google says the first set of to-dos and topics can take a few hours after smart features are enabled and, in some cases, as long as seven days. A quiet account may show an empty AI Inbox.

What AI Inbox includes—and what it misses

AI Inbox currently focuses on Gmail’s Primary tab. It does not use messages filed under Social or Promotions, and it filters out spam plus messages that are archived, deleted, muted or snoozed. That means a valid deadline can be absent simply because the message sits outside Primary.

Google’s help documentation also says AI Inbox does not support attachments, delegated inboxes or encrypted email. A task described only in a PDF, a shared mailbox or a client-side encrypted message may therefore be invisible to the feature.

The practical rule is simple: treat AI Inbox as a triage layer, not a complete task manager. Continue checking labels, calendar events, shared mailboxes and any project system that holds obligations outside ordinary Gmail messages.

Availability and rollout timing

Gmail Live began a gradual rollout to Rapid Release and Scheduled Release domains on October 15, with Google allowing up to 15 days for visibility. Eligible editions include Business and Enterprise plans, Frontline Plus, Workspace for Nonprofits, several Gemini add-ons, Google AI Pro for Education and Google AI Plus, Pro and Ultra consumer plans.

AI Inbox started reaching Rapid Release domains on October 12 and is scheduled to begin for Scheduled Release domains on October 26, also with a gradual rollout of up to 15 days. Google lists Business, Enterprise, Frontline Plus and AI Pro for Education availability, but separately notes that Starter and Education editions will receive the feature later. That means plan eligibility does not guarantee the same launch date.

Both features currently require a U.S. account experience with English set to EN-US. App version, administrator settings, account type and rollout stage can all explain why one eligible user sees the new menu while another does not.

Privacy and history controls deserve a review

Gmail Live saves transcripts in the user’s Gemini in Workspace conversation history. Google says users can review and delete that history from a web browser. The microphone can be muted during a session, and Gmail cannot start Live if microphone permission is denied.

AI Inbox generates its recommendations from content and activity across Workspace apps. Google notes that some to-dos or topics may be generated for optimization without appearing in the visible AI Inbox. Users can open View all on the web to inspect more of that generated data.

For work accounts, administrators control several Gemini and smart-feature settings, but Google says AI Inbox itself currently has no dedicated admin control. Organizations handling regulated or confidential information should review Workspace’s data-protection terms, retention policies and beta-feature settings before treating either feature as a standard workflow.

Five checks before relying on Gmail’s AI features

• Open the cited email and confirm the latest date, sender and thread before acting.
• Check attachments manually because AI Inbox does not support them.
• Look outside Primary for deadlines that may be in Promotions, an archived label or a delegated inbox.
• Review Gemini in Workspace conversation history and delete sensitive transcripts when appropriate.
• Keep important tasks in a real calendar or project system instead of assuming AI Inbox is complete.

A practical device for Gmail Live

Gmail Live is built for mobile use, so a current Android phone with strong battery life and reliable microphones is the most natural fit. Readers comparing compatible hardware can view the unlocked Google Pixel 11 on Amazon.

ReaderSpice may earn a commission from qualifying purchases made through affiliate links, at no additional cost to you.

Bottom line

Gmail Live and AI Inbox make Gmail more proactive in two different ways: one answers spoken questions with source cards, while the other organizes tasks and priority topics before the user searches. For busy inboxes, both can reduce the time spent opening threads.

Their limits matter as much as their convenience. U.S.-only availability, English requirements, beta status, rollout delays and unsupported message types mean neither feature should be treated as a complete record. The safest workflow is to use Gemini for triage, then verify every consequential detail in the original email.

Sources

Google Workspace Updates: Gmail Live rollout for Workspace users

Gmail Help: Talk naturally with Gmail Live

Google Workspace Updates: Gmail AI Inbox rollout

Gmail Help: Manage to-dos and topics with AI Inbox

Google: Gmail is entering the Gemini era