The Google-Native Agent That Can See and Act
Google shipped a new frontier model on September 30 and its own API doesn’t list it yet. Gemini 4 Argon sits above the whole Flash line at an introductory $2 per 1M input tokens and $10 per 1M output, with cached input 95 percent off. No models-page entry. No pricing row. No model card. Second thing, and it costs money. Google has put a May 7, 2027 shutdown date on gemini-3.1-flash-lite, the cheapest model it sells, and named the dearer gemini-3.5-flash-lite as the replacement. Gemini 3.8 Flash is still the newest Flash model and Gemini 3.1 Pro is still a preview. Name the pricing tier every time, because batch is exactly half of standard.
One intelligence layer across Google’s working surfaces
Gemini isn’t one chat product. It’s a family of models and experiences: Search, the Gemini app, Workspace, Android, developer tools, voice and creative systems. Sold through consumer plans as well as the API.
Fast, multimodal and connected
Gemini is strongest when the work mixes current information, visual or voice or audio context, and a Google service that can help finish the job.
- Search and Maps grounding
- Text, image, voice and video
- Consumer, enterprise and developer surfaces
Information-rich action
Use Gemini for current research, Workspace creation, Android assistance, live voice agents, proactive monitoring and agents that need to see and operate interfaces.
- Research and generative UI
- Gmail, Docs, Slides and Sheets
- Browser, mobile and desktop automation
The context already lives with Google
Gemini has a built-in head start when the task begins in Search, Gmail, Drive, Android, Maps, YouTube or a Google developer environment.
- Less context transfer
- Live information
- One model family across many surfaces
A new frontier model with no price page, and a cheap model Google is retiring in favor of a dearer one
Name the tier or the number means nothing. Every rate quoted here is paid-tier standard. Batch is exactly half of it. Now the lineup, which moved twice in two days. On September 30, 2026 Google announced Gemini 4 Argon, a frontier model above the whole Flash line, at an introductory $2 per 1M input and $10 per 1M output. It has no entry on the API models page, no row on the pricing page and no model card. Gemini 3.8 Flash, generally available since September 2, is still the newest Flash model. Gemini 3.1 Pro is still a preview endpoint, seven and a half months in. And the deprecation calendar isn’t empty any more. gemini-3.1-flash-lite now has a May 7, 2027 shutdown date. The May 2026 Antigravity agent stops on October 5. And gemini-2.5-flash-image stops October 2, with a named replacement the catalog doesn’t list.
gemini-2.5-flash-image; the replacement ID Google names is absent from the models pagegemini-3.1-flash-lite; the dearer gemini-3.5-flash-lite is the named replacementantigravity-preview-05-2026, replaced by antigravity-preview-09-2026Gemini 4 Argon Announced September 30
A new frontier model, and you can’t buy it from the API yet. Both halves matter. Google says Argon “delivers frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense.” It prices it at an introductory $2 per 1M input tokens and $10 per 1M output, with cached input tokens at 95 percent off. Google also raised the output ceiling, “significantly expanding the model’s output token limit to an industry-leading 1M tokens,” up from 64K. Access runs in stages. First “a set of trusted cyber defenders through our Fairwind Program,” then developers, enterprises and consumers, “starting with paid API customers and Google AI Ultra subscribers.” Now the other half. Argon has no entry on the API models page, no row on the API pricing page and no model card. So the context window, the knowledge cutoff and the API model ID are all unpublished, and the only dated source for the price is the announcement itself.
· Google’s wording, September 30, 2026: “Today, we’re announcing our new frontier model, Gemini 4 Argon, which is rolling out to a set of trusted cyber defenders through our Fairwind Program.”
- Do not put Argon in a build plan yet: there is no model ID to pin
- The $2 and $10 figures are introductory and announcement-sourced, not pricing-page rates
- Ultra subscribers and paid API customers are named first in the rollout order
Gemini 3.8 Flash GA September 2
Still the newest Flash model, no longer the newest model. Its DeepMind model card gives up to 1M tokens of context, 64K tokens of output and a March 2026 knowledge cutoff. Google’s framing is enthusiastic, and it’s Google’s. The launch post calls it “our most intelligent workhorse model, delivering significant improvements from 3.7 Flash across software engineering, agentic tasks, and critical, multi-step reasoning.” The docs call it “Our most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows.” Gemini 4 Argon now sits above it, so “the newest Gemini model” and “the newest Flash model” are two different answers.
- Pin
gemini-3.8-flash - Knowledge cutoff March 2026, published on the model card
- Same promotional pricing as 3.7 and 3.6 Flash, including the January 1, 2027 doubling
- It is the Flash flagship, not the lineup flagship
The knowledge cutoff Google publishes in one place and omits in the other Check both pages
Nobody re-checks a negative. That’s how a missing field turns into a false claim. The Gemini 3.8 Flash model card, dated September 2, 2026, publishes a March 2026 knowledge cutoff. Its caveat: in some domains users “may experience the model’s knowledge is limited to January 2025.” Google’s API model page lists the 1,048,576-token input limit, the 65,536-token output limit and a September 2026 update stamp, and it simply has no cutoff field. So read both pages before you report that Google hasn’t published something. The card is a dated specification. The API page just leaves the field out, so there’s no conflict to resolve. Quote the card. An omission isn’t a denial.
- State March 2026, with the January 2025 caveat attached
- Two Google surfaces, one figure and one blank: quote the card
- Where the vendor contradicts itself, say which document you trusted and why
Gemini 3.1 Pro The shipping Pro model
gemini-3.1-pro-preview is what a buyer actually gets when they ask for Pro. Note the suffix. Google launched it on February 19, 2026 with the words “we are releasing 3.1 Pro in preview today to validate these updates,” and promised general availability soon. Seven and a half months later it’s still a preview endpoint. Google describes it as offering “advanced intelligence, complex problem-solving skills, and powerful agentic and vibe coding capabilities.” It’s priced on the paid tier, and it’s the Pro model inside the consumer AI plans. Its pricing is context-tiered, and both tiers are published. Paid-tier standard is $2.00 per 1M input and $12.00 per 1M output for prompts of 200k tokens or fewer, and $4.00 and $18.00 above 200k. Batch halves both: $1.00 and $6.00 on the lower tier, $2.00 and $9.00 on the upper one. Cost the tier your prompts will actually land in, because a long-context pipeline is on the second rate, not the first.
- Pin
gemini-3.1-pro-previewand price the context tier you will actually send - In preview since February 19, 2026, which is a second slipped Pro timeline
- Do not tell a client that Gemini has no Pro model; tell them the Pro model is a preview
Gemini 3.1 Deep Think Sold, not documented
Google sells Deep Think and documents almost nothing about it. The only current first-party trace is a feature line inside the top consumer plan, “higher access to Deep Think reasoning mode.” That names a capability. It doesn’t name a model version, a context window, a price or an API endpoint, and DeepMind no longer lists a Gemini 3.1 Deep Think entry at all. So you can tell a client the Ultra plan includes a Deep Think mode. You can’t tell them which model it runs, and you should not put a version number in a proposal.
- Bundled with the highest consumer tier
- No version number is currently sourceable
- Confirm availability and plan gating in the product before you quote it
Gemini 3.8 Live and 3.8 Live Extended Thinking GA September 15
Two audio-to-audio Live API models, both generally available and both listed as stable. Google positions gemini-3.8-live as the “default Live API model for most low-latency voice agent experiences without reasoning delays,” and gemini-3.8-live-extended-thinking as the “high-reasoning Live API model for voice interactions.” Both are available through the Gemini API and Google AI Studio. The low-latency model is also on Search Live. Enterprise access is a private preview. Google says the pair “deliver the building blocks for reliable, production-ready voice agents,” which is a vendor claim. Paid-tier standard is $0.75 per 1M input and $4.50 per 1M output for text, with the free tier free of charge. Extended Thinking is rolling out to Workspace and Gemini Live subscribers in Docs, Gmail and Keep.
- Choose latency or reasoning, then pin the matching ID
- Google recommends migrating off
gemini-3.1-flash-live-preview, now marked legacy - Enterprise deployment is private preview, so do not promise it
Gemini 3.7 Flash Previous generation
Generally available August 13, superseded September 2. The docs now describe it as “Our previous-generation Flash model for complex coding, agentic workflows, and reliable multi-step execution.” Its model card gives a 1,048,576-token input limit, a 65,536-token output limit and a March 2026 knowledge cutoff. Its pricing is identical to 3.8 Flash.
- Pin
gemini-3.7-flashif you are staying on it - No forced migration to 3.8 Flash has been announced
- No shutdown date has been published for it
- Same promotional rate, same January 1 step-up
The recycled superlative Read this before you quote Google
Three Flash generations. One adjective. At its August 13 launch Google called Gemini 3.7 Flash “our most intelligent workhorse model yet for coding and agents.” Three weeks later near-identical framing belongs to 3.8 Flash. The words were accurate on both days and distinguishing on neither. So treat a vendor superlative as a marketing constant. Take the differences from the model card, the price table and your own traces.
- The same superlative has now covered three Flash generations
- Quote it with its date attached or not at all
- Benchmark the models yourself rather than reading the adjectives
Gemini 3.6 Flash Two generations back
The July 21 default, still listed and still on exactly the same promotional pricing as 3.7 and 3.8 Flash. The models page has re-described it as “Our previous-generation Flash model, balancing speed and multimodal capabilities across general agentic and everyday tasks.” Its distinguishing claim travels with it. Google wrote that “according to the Artificial Analysis Index, it reduces output token usage by 17% compared to 3.5 Flash.”
- Pin
gemini-3.6-flashif you are staying on it - The 17 percent figure is the Artificial Analysis Index, cited by Google, not a Google benchmark
- No shutdown date has been published for it
- Same promotional rate, same January 1 step-up
Gemini 3.1 Flash-Lite Shuts down May 7, 2027
Google has put a shutdown date on the cheapest model it sells. gemini-3.1-flash-lite runs at $0.25 per 1M input tokens for text, image and video and $1.50 per 1M output. The deprecations page now schedules it to stop serving on May 7, 2027. The named replacement is gemini-3.5-flash-lite, at $0.30 and $2.50. That’s 20 percent more on input and 67 percent more on output. Meanwhile the models page still sells 3.1 Flash-Lite as “frontier-class performance rivaling larger models at a fraction of the cost,” with no deprecation mark on it. So the catalog and the schedule tell a buyer two different things about the same model on the same day.
- Any high-volume pipeline pinned here has a migration with a price increase attached
- Google publishes May 7, 2027 as the earliest possible date, so treat it as a planning floor
- Price the replacement now: you have about seven months before the published date
Gemini 3.5 Flash-Lite Now the named replacement
The dearer Flash-Lite is the one with a future. List price is $0.30 per 1M input tokens and $2.50 per 1M output on the paid standard tier. Google reports roughly 350 output tokens per second. And it’s the model Google names as the replacement when 3.1 Flash-Lite shuts down on May 7, 2027. It carries no shutdown date of its own. Google also tells new projects to start here or on 3.8 Flash rather than on the 2.5 family.
- Pin
gemini-3.5-flash-litefor new high-volume work - Budget the step up from $0.25 / $1.50 if you are moving off 3.1 Flash-Lite
- No shutdown date has been published for it
Gemini 3.5 Flash Now described as legacy
The May 19 model has survived three newer defaults, but Google’s language about it has changed. The models page now calls it “our legacy Flash model, providing baseline speed and foundational performance for routine, high-throughput workloads.” 3.6 Flash gets “previous-generation.” And it sits outside the promotion at $1.50 per 1M input and $9.00 per 1M output, with caching at $0.15 and batch at $0.75 and $4.50. So the oldest supported Flash is the expensive one.
- “Legacy” is Google signaling the next deprecation candidate, not a shutdown date
- No forced migration has been announced
- You are paying a premium to stay on it: move deliberately, but do plan the move
Gemini 3.5 Pro Google has stopped saying it is coming
A promise can expire without anybody canceling it. This is what that looks like. On May 19, 2026 Google said Pro was “already being used internally, and we look forward to rolling it out next month.” The month it named was June. It’s now October, and the model is still absent from the models page and the pricing page. On July 21 Google said only that it’s testing with partners and will be broadly available when ready. Since then the last surface carrying a forward-looking line about it has been rebuilt around Gemini 4 Argon, and no current DeepMind page mentions Gemini 3.5 Pro. So the honest status isn’t “coming soon.” It’s that Google has stopped saying it’s coming, and has published nothing canceling it either.
- The month Google named was June; it is now October
- Do not promise availability or a date, and do not repeat “coming soon”
- Gemini 3.1 Pro is the Pro model that exists today
Computer use Rebuilt around three environments
The computer-use documentation no longer publishes a supported-model list. It now says only that you should check Model versions for the list, and organizes everything as Gemini 3.x against Gemini 2.5 legacy, with gemini-3.8-flash, gemini-3.5-flash and gemini-2.5-computer-use-preview-10-2025 appearing in code samples. What the page gained is bigger than what it lost. It documents three target environments, browser, mobile and desktop, plus thinking-level control, configurable safety policies, safety overrides and first-party prompt injection detection for Gemini 3.x. That last one is a real control, not a warning. Turn it on.
- Verify computer-use support against the model ID you pin, in the product, not from a list
- Three environments: browser, mobile and desktop
- Prompt injection detection is now a Google feature, so enable it rather than writing your own filter
Gemini 3.8 Flash Cyber Fairwind only
The cyber model moved up three generations, from Gemini 3.5 Flash Cyber to Gemini 3.8 Flash Cyber, and changed its access route on September 2, 2026. It’s “only available to trusted defenders who require a more comprehensive set of cyber capabilities.” Access now runs through the Fairwind Program, “a limited access program for governments and trusted partners to use our cyber defense tools.” CodeMender remains the tooling layer. Google says “CodeMender with Gemini 3.8 Flash Cyber delivers the specialized reasoning to write and validate code fixes, at a fraction of the operating cost.” Google Cloud customers outside Fairwind can still use CodeMender with publicly available models through the Gemini Enterprise Agent Platform. Gemini 4 Argon enters through the same program, and DeepMind’s Argon cyber page benchmarks it against 3.8 Flash Cyber, which is the clearest signal of where this line is heading.
- Not a purchasable capability
- The Fairwind page does not mention Gemini 3.5 Flash Cyber at all, so its fate is unconfirmed rather than concluded
- CodeMender is available more widely than the cyber model is
Two new text-to-speech models Stable September 22
Voice output stopped being a preview. First, gemini-3.8-flash-tts. It’s the “flagship creative text-to-speech model for studio-grade voice fidelity, expressive acting, Voice design, and Voice replication.” gemini-3.8-flash-lite-tts is the “fast, cost-efficient text-to-speech model for high-volume production, real-time voice agent cascades, and Voice replication.” Both arrive with a Voices endpoint and an extended library Google describes as 150 or more prebuilt and custom voices. The old preview ID, gemini-3.1-flash-tts-preview, is now marked legacy. Watch the price shape. Paid-tier standard is $0.50 per 1M input and $9.00 per 1M output through December 31, 2026, then $1.00 and $18.00. That’s a doubling on the same day the Flash text promotion ends.
· Both IDs and the September 22, 2026 general-availability date come from Google’s API release notes, which Google rewrites in place. Confirm them directly before you commit.
- Pin the explicit TTS ID, and say which one in the contract
- Price the 2027 rate, not the promotional one, for anything running past December 31
- Voice replication needs written consent from the person whose voice it is
| Model or capability | Paid-tier standard, per 1M tokens | What a buyer needs to know |
|---|---|---|
| Gemini 4 Argon | $2.00 in / $10.00 out, introductory | From the September 30, 2026 announcement only. There is no Argon row on the pricing page and no Argon entry on the models page, so there is no API model ID to quote. Cached input is 95 percent off input. |
gemini-3.8-flash, gemini-3.7-flash, gemini-3.6-flash | $0.75 in / $3.75 out through December 31, 2026, then $1.50 / $7.50 | Batch is $0.375 / $1.875 now and $0.75 / $3.75 from January 1. Free tier is free of charge. |
gemini-3.1-pro-preview | $2.00 in / $12.00 out at 200k tokens or fewer; $4.00 in / $18.00 out above 200k | Context-tiered, so a long prompt is on the second rate. Batch halves both tiers: $1.00 / $6.00 and $2.00 / $9.00. |
gemini-3.5-flash | $1.50 in / $9.00 out | Outside the promotion, so the model Google now calls legacy is the expensive one. Caching $0.15, batch $0.75 / $4.50. |
gemini-3.1-flash-lite | $0.25 in / $1.50 out | Cheapest on both sides, and scheduled to shut down on May 7, 2027. Input rate covers text, image and video. |
gemini-3.5-flash-lite | $0.30 in / $2.50 out | The replacement Google names for 3.1 Flash-Lite, at 20 percent more on input and 67 percent more on output. |
gemini-3.8-flash-tts, gemini-3.8-flash-lite-tts | $0.50 in / $9.00 out through December 31, 2026, then $1.00 / $18.00 | Batch is $0.25 / $4.50 now and $0.50 / $9.00 from January 1. The output rate doubles on the same date as the Flash text promotion ends. |
gemini-3.8-live and gemini-3.8-live-extended-thinking | $0.75 in / $4.50 out for text | Free tier free of charge. Enterprise access is private preview. |
gemini-omni-1.1-flash | $1.50 in for text, image, video and audio / $9.00 out for text | No free tier at all, so a free-tier assumption carried over from the Flash text models fails here. |
gemini-3.5-transcribe | $2.00 or $0.003 per audio minute in / $12.00 or $0.002 per minute out | Batch transcription is the cheaper of the two IDs by a wide margin. |
gemini-3.5-transcribe-live | $3.50 or $0.005 per audio minute in / $21.00 or $0.004 per minute out | 75 percent more per 1M tokens than the batch model, and per audio minute 67 percent more on input and 100 percent more on output. Pick the ID on latency need, not habit. |
gemini-3.1-flash-image (Nano Banana 2) | $0.50 in for text and image / output is $3.00 for text and thinking plus $60.00 for images | Output has two components, not one. Google states the image rate as equivalent to $0.045 per 0.5K image. Batch is $0.25 in / $1.50 text plus $30.00 images. No free tier. |
gemini-3.1-flash-lite-image (Nano Banana 2 Lite) | $0.25 in / $30.00 out for images | The cheap image tier is still expensive next to text. |
gemini-3-pro-image (Nano Banana Pro) | No rate published | Listed as stable on the models page with no section on the pricing page, so there is no first-party rate to quote. Get a number in writing before you build an image pipeline on it. |
| Search grounding | 5,000 free requests per month, then $14 per 1,000 | The free allowance is shared across all Gemini 3.x models, so it does not multiply when you add one. |
| Shutdown date | Model | Replacement named by Google |
|---|---|---|
| August 31, 2026 (passed, unconfirmed at thirty-one days) | gemini-robotics-er-1.6-preview | gemini-robotics-er-2-preview, though three Google pages still present the 1.6 preview as live |
| September 30, 2026 (passed; row deleted rather than marked shut down) | gemini-omni-flash-preview | gemini-omni-1.1-flash. The pricing page still sells the retired endpoint |
| October 2, 2026 (scheduled, as read October 1) | gemini-2.5-flash-image | gemini-3.1-flash-image-preview in the deprecation table, an ID that appears nowhere on the models page. The models page lists gemini-3.1-flash-image as stable, and still lists gemini-2.5-flash-image as stable with no deprecation mark |
| October 5, 2026 | antigravity-preview-05-2026 | antigravity-preview-09-2026, which uses a different tool contract |
| May 7, 2027 | gemini-3.1-flash-lite | gemini-3.5-flash-lite, which costs more on both input and output |
Gemini’s advantage is the stack around the model
Each surface brings a different kind of context. Search knows the live web. Workspace knows the work. Android knows the moment. And the API turns the model into a building block. September closed with two changes a buyer has to act on. Skills replaced Gems, and Gems now has deprecation dates attached to it. And Gemini 3.8 Live with Live Avatar shipped into Gemini Enterprise, which is the first Live capability to reach an enterprise product rather than a private preview. Meanwhile the consumer Gemini Drop has missed two consecutive months outright. This is also where a buyer meets the consumer plans, which decide which model an individual account gets, and how much of it.
Gemini app
Conversation, connected apps, multimodal input, Daily Brief and emerging personal agents. What the account can reach depends on its AI plan.
Search AI Mode
Live web grounding, information agents, Maps and custom generative interfaces.
Google Workspace
Gmail, Docs, Slides, Sheets, Meet and organizational content, with Gemini 3.8 Live Extended Thinking rolling out to Workspace and Gemini Live subscribers in Docs, Gmail and Keep.
Gemini Notebook
Source-grounded notebooks, audio and video overviews, reports and study tools. Renamed from NotebookLM on July 16, 2026, with a secure cloud computer for code execution and sync with the Gemini app.
Gemini API
Model access, built-in tools, computer use, embeddings and application-specific agents.
Antigravity
Agent-first development with artifacts, plans, screenshots and execution evidence. The default agent changed on September 17, 2026.
Android
Camera, voice, notifications, connected apps and contextual assistance close to the moment of work. At Galaxy Unpacked on July 22, 2026 Google said Gemini Intelligence task automation now reaches more than 40 apps. Gemini Notebook comes preinstalled on the Galaxy Z Fold8, and Gemini runs on the Galaxy Watch 9.
Omni, Veo, Lyria and image tools
Image, video and music creation connected to Gemini reasoning and the wider Google ecosystem.
Skills replaces Gems, and Gems gets an end date
A feature your users built things in is being retired, in stages. Skills are invoked by typing a forward slash and the skill name in the prompt bar, they stack, and they take reference files including text, PDFs and images. They roll out to every Google AI subscription tier in Gemini chat globally for users 18 and over, with Workspace following. Here is the part for an administrator. Gems shut down in November 2026 for personal accounts, March 2027 for Workspace business, enterprise and nonprofit, and June 2027 for Workspace education. That’s a user-facing deprecation with three dates, so inventory who in your organization has built a Gem before November arrives.
Gemini 3.8 Live with Live Avatar, in Gemini Enterprise
Google’s words: “Starting today, Gemini 3.8 Live with Live Avatar is available in Gemini Enterprise.” Real-time animated avatars with lip-sync across 97 languages, asynchronous tool execution so the avatar keeps talking while a tool runs, and SynthID watermarking on the output. This is the first Live capability to land inside an enterprise product rather than a private preview, so the honest line about enterprise Live access has changed. Avatar is available in Gemini Enterprise. Direct enterprise access to the Live API models is still private preview. Those are two different purchases.
A new Antigravity agent, and the old one stops October 5
Google’s API release notes record antigravity-preview-09-2026 replacing the May 2026 version. Built-in tools moved to PascalCase parameters, and file edits became line-range replacements instead of full rewrites. Both are breaking changes for any integration written against the old contract. And there’s now a date on the old agent. antigravity-preview-05-2026 shuts down on October 5, 2026. This item appears only in the release notes and the deprecations page, which Google rewrites in place. No dated article on Google’s blogs, DeepMind or the developer blog carries it. So confirm it directly, then move. Eighteen days from release to shutdown isn’t a migration window.
Gemini 3.8 Live reached GA
Two audio-to-audio models, gemini-3.8-live for low latency and gemini-3.8-live-extended-thinking for reasoning, generally available through the Gemini API and Google AI Studio. The low-latency model is also on Search Live. Enterprises get private preview. The same release marks gemini-3.1-flash-live-preview as a legacy Live API preview and recommends moving to 3.8 Live.
Lyria 3.5 reached GA
A full-length song generation model with, in Google’s words, “improved musical coherence, natural vocals, and fine-grained duration and structural control,” generating 44.1 kHz stereo audio from text and image inputs. It shipped into the Gemini app. But the API models page, read October 1, 2026, lists only Lyria 3 IDs, so there’s no Lyria 3.5 model ID to pin from it. One caution on the date. September 3 comes from Google’s API release notes, and the announcement page didn’t surface a publication date when it was read. Confirm the date directly rather than quoting it as settled.
Agentic video understanding
Google introduced agentic video understanding for Gemini 3.7 Flash, Gemini 3.6 Flash and Gemini 3.5 Flash-Lite. Google claims up to 88% fewer tokens, 66% lower cost and 7% better accuracy, at no extra charge. Those are Google’s figures from Google’s own announcement, not an independent measurement. Quote them as vendor claims, and test them on your own footage before you rebuild a pipeline around the savings. The flagship 3.8 Flash is still not named, and nothing published since revisits that.
Gemini Omni Flash
The model ID is gemini-omni-1.1-flash. It adds video extension, interpolation and configurable resolution at 360p, 720p, 1080p and 4k. It’s priced on the paid standard tier at $1.50 per 1M input and $9.00 per 1M output, with no free tier at all. Google’s own pages give four different answers about its status. The models page lists it under a preview heading. DeepMind lists it as GA. The release notes headline says GA. And the dated announcement says neither, calling it “production-ready” and “rolling out across the Google developer ecosystem.” The preview endpoint it replaces, gemini-omni-flash-preview, was scheduled to stop serving on September 30, 2026. That date has passed, its deprecation row has been deleted, and the pricing page still sells it.
Gemini 3.5 Transcribe
Two model IDs, gemini-3.5-transcribe and gemini-3.5-transcribe-live, with support for 85 or more languages, speaker diarization and custom vocabulary biasing. The live model costs 75 percent more per 1M tokens than the batch model, and between 67 and 100 percent more per audio minute depending on direction. So the choice between them is a budget decision as well as a latency one.
Gemini Embedding 2
Google calls Gemini Embedding 2 its “first multimodal embedding model, mapping text, images, video, audio, and PDFs into a unified embedding space.” Check the ID before you build on it. The API models page, read October 1, 2026, lists gemini-embedding-001 as stable and has no Gemini Embedding 2 entry. DeepMind’s model index lists Gemini Embedding with no version. A unified space across five modalities is a real retrieval capability. But an ID you can’t find in the catalog isn’t one you can pin.
The August Workspace Feature Drop
Screenshots in Meet notes, automatic organization of meeting artifacts in Drive, quiz generation in Forms, Workspace Studio skills invoked by @mention, and enterprise security controls in Studio Flows. Workspace features arrive by edition and admin policy, so confirm eligibility before you describe any of these as available to your organization. As of October 1, 2026 it’s the newest Workspace Feature Drop. But several newer Workspace posts sit above it, so “most recent drop” and “most recent Workspace news” aren’t the same thing.
Two consecutive months with no Gemini Drop
As of October 1, 2026, the July 31 drop is the most recent one. Neither an August nor a September edition exists, and September’s window closed on September 30. The drops have migrated between /products-and-platforms/products/gemini/ and /innovation-and-ai/products/gemini-app/, so both paths matter. Both 404 for both months, and the Gemini app blog index lists no successor. So the ordinary explanation, that the page simply moved, is ruled out. October isn’t due yet: the July edition published on the 31st, so the test is October 31. Two missed months is a gap. Three would be a pattern.
New connected apps and services
Google added Granola, Otter.ai, Wix, OpenTable (UK), Ticketmaster, Zocdoc and others to the Gemini app’s connected apps on August 12, 2026. Every connector widens what the assistant may read and act on. So treat the list as a permissions decision, not a feature announcement.
| Surface | Context it owns | Best outcome | Availability caution |
|---|---|---|---|
| Gemini app | Personal conversation and connected apps | Everyday help and proactive assistance | Features vary by plan, language and region; the AI plan decides which Pro model and how much of it the account reaches |
| Search AI Mode | Live web, shopping and Maps | Grounded answer or interactive interface | Generative UI and agents roll out over time |
| Workspace | Organizational mail, files and meetings | Native work artifact | Edition and admin policy matter; the August 26 feature drop was the most recent drop as of October 1, 2026, and Gems shut down for Workspace business, enterprise and nonprofit in March 2027 and for education in June 2027 |
| Voice and Live | Spoken conversation in real time | Production voice agents and hands-free help | Generally available through the API and AI Studio since September 15, 2026; Gemini 3.8 Live with Live Avatar shipped in Gemini Enterprise on September 24, 2026, while direct enterprise access to the Live API models remains private preview |
| Gemini Notebook | Curated user source set | Source-grounded learning and briefing | The cloud computer is on AI Ultra and Workspace Expanded Access; Google’s July 16, 2026 post promised Pro on the web in coming weeks, not confirmed as shipped as of October 1, 2026 |
| API / Antigravity | Developer-defined tools and environment | Custom agent or software workflow | Requires safety and permission design; the default Antigravity agent and its tool contract changed on September 17, 2026 |
| Android | Device, camera and contextual moment | Personal assistance and action | Hardware and rollout vary |
| Plan | What Google lists in it | Price, with the date Google stated it |
|---|---|---|
| Google AI Plus | 400 GB storage, 2x access to Gemini, Gemini 3.1 Pro, Deep Research, image, music and video generation, Gemini Notebook, family sharing | $7.99 a month in the US, stated January 27, 2026. Google adds that price varies by country, and the plans page renders the current amount client-side. |
| Google AI Pro | 5 TB storage, 4x access to Gemini, expanded Gemini 3.1 Pro, Deep Research, Gemini Spark, Google Flow, $10 a month in Google Cloud credits, Google Home Premium Standard | $19.99 a month, stated August 19, 2026 in the student-offer post. |
| Google AI Ultra | From 20 TB storage, up to 20x access to Gemini, highest Gemini 3.1 Pro limits, Deep Think, Project Genie, YouTube Premium individual, $40 a month in Google Cloud credits, Google Home Premium Advanced | $200 a month for the top tier, reduced from $250, plus a $100 tier for developers, technical leads, knowledge workers and advanced creators. Both stated May 19, 2026. Google has not mapped the storage and access figures to a specific tier. |
Google is turning Search and the device into persistent workers
The Gemini 3.5 through 3.8 line is built to sustain work, not just answer prompts. Spark, background information agents, voice agents and computer use are Google trying to make intelligence persistent across time and interfaces. Most of those surfaces are still limited, regionally excluded or unconfirmed. September added two first-party controls worth having: an oversight layer for running agents, and prompt injection detection built into computer use.
Gemini Spark Not in the EEA, Nigeria, Switzerland or the UK
A personal agent meant to help across the digital day. The July Gemini Drop framed the release as “Gemini Spark is going global... now available worldwide.” Google’s support page states the real list: “Available wherever Gemini Apps are supported, except in the European Economic Area, Nigeria, Switzerland and the United Kingdom.” It’s also gated to Google AI Pro and above. So read the exclusion list and the plan requirement, not the headline.
- “Worldwide” with four named exclusions, unchanged for two months
- Unavailable to readers in the UK, the EEA, Switzerland and Nigeria
- Inside supported regions it still needs a Pro or Ultra plan
Computer use Prompt injection detection shipped
The computer-use documentation was rebuilt, and the most useful thing in it is a defense. Google now documents prompt injection detection for Gemini 3.x computer use, alongside configurable safety policies, safety overrides and thinking-level control, across three target environments: browser, mobile and desktop. The capability launched on Gemini 3.5 Flash on June 24, 2026 and has been extended since. One practical consequence of the rebuild. The page no longer publishes a supported-model list. It defers to the model versions reference. So there’s no page you can point at to prove a given Flash model is or isn’t supported. Test the ID you intend to pin.
- Turn on prompt injection detection rather than writing your own filter
- Browser, mobile and desktop are separate environments with separate risk
- No published model list any more: verify support against the ID you pin, every time
Managed agents and the new Antigravity default Old agent stops October 5
A model swap shows up in a dashboard. A tool-contract change shows up as failing agents. This is the second kind, and it now has a deadline. Google opened managed agents to free-tier projects on July 28, 2026: “Developers can experiment with agentic workflows using an API key from a project without active billing.” That post named antigravity-preview-05-2026 as running Gemini 3.6 Flash by default. Google’s API release notes, dated September 17, 2026, record antigravity-preview-09-2026 replacing it, with built-in tools moving to PascalCase parameters and file replacements expressed as line ranges rather than full rewrites. Both are breaking changes for any integration written against the old contract. And the deprecations page shuts the May 2026 agent down on October 5, 2026. If something you own still pins the old ID, move it before that date.
antigravity-preview-05-2026stops serving October 5, 2026- PascalCase tool parameters and line-range file edits are breaking changes
- Release-notes and deprecations-page evidence only: verify on the pages before you act
Watching the agent while it runs Private preview
Until last month the safe computer-use workflow above had no first-party Google control plane attached to it. Two posts on Google’s developer blog change that. On September 16, 2026 Google described Agent Anomaly Detection, now in private preview on the Gemini Enterprise Agent Platform. It’s an oversight layer that analyzes OpenTelemetry traces to catch behavioral risks without affecting live request performance. On September 15 Google published guidance on building zero-trust AI agents that judge intent rather than syntax, which is runtime governance rather than prompt hygiene. Read both before an agent touches production. Neither is generally available, which makes the prompt injection detection now documented for computer use the one control you can actually turn on today.
- Trace-based oversight rather than prompt-level filtering
- Private preview, so do not scope a program around it yet
- Prompt injection detection is the shipped alternative while you wait
Voice agents Avatar shipped September 24
Gemini 3.8 Live and 3.8 Live Extended Thinking make production voice agents a supported build rather than a preview experiment. The low-latency model is aimed at conversational turn-taking. The extended-thinking model is for problems that need reasoning mid-call. And on September 24, 2026 Google put a face on it: “Starting today, Gemini 3.8 Live with Live Avatar is available in Gemini Enterprise.” That means lip-synced animated avatars across 97 languages, asynchronous tool execution and SynthID watermarking. Note which product that is. Avatar ships inside Gemini Enterprise. Direct enterprise access to the Live API models is still private preview.
- Latency and reasoning are separate model choices
- Live Avatar is a Gemini Enterprise feature, not a Live API entitlement
- Design the escalation path to a human before you launch
Agentic video understanding September 1
Google’s agentic video capability applies to Gemini 3.7 Flash, Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, and Google says it costs nothing extra. Google claims up to 88% fewer tokens, 66% lower cost and 7% better accuracy. Note which models are named. The flagship 3.8 Flash isn’t among them, and Google has published nothing since to explain the exclusion.
- Vendor figures, attributed to Google
- Named for 3.7 Flash, 3.6 Flash and 3.5 Flash-Lite
- Validate the savings on your own video before repricing a pipeline
Information agents in Search
Agents can watch an ongoing information need and report what changed, with links for deeper action.
- Repeated research
- Alerts and monitoring
- Source review remains necessary
Daily Brief
The Gemini app can pull timely personal information into one proactive morning view.
- Reduce manual checking
- Depends on connected context
- Review privacy and relevance
Generative interfaces
Search can build a purpose-built interface or simulation instead of forcing every answer into prose. Google promised this for everyone in Search “this summer” on May 19, 2026. The summer it named ended on September 22, and no post confirms that it shipped. So the window closed with the promise unmet and unexplained.
- Promised, the window has closed, still unconfirmed
- Interactive decision support
- Generated logic still needs testing
Android Halo
A status-bar surface that shows agent activity, announced May 19, 2026 in preview and described as rolling out later this year. It’s a visibility surface, not a management console. Nothing further had been published as of October 1, 2026.
- Agent activity, not agent control
- Announced in preview
- Rollout status matters
Gemini is joining reasoning and media production
Image output is charged at rates two orders of magnitude above text, and the image models have no free tier at all. That’s the least intuitive pricing in the product, so start there. What the creative stack is good for is connecting a source-rich research process straight to image, video, music and presentation outputs, without shrinking creativity down to a single model name.
Nano Banana family Three IDs, two published prices
Google’s image stack is three stable models with very different economics, all on the paid tier and none with a free tier. Nano Banana 2 is gemini-3.1-flash-image at $0.50 input for text and image. Its output has two parts, and the second one gets missed: $3.00 for text and thinking plus $60.00 for images, which Google describes as equivalent to $0.045 per 0.5K image. Cost a pipeline on the $60 alone and you’ve left a charge out. Nano Banana 2 Lite is gemini-3.1-flash-lite-image at $0.25 and $30.00. Nano Banana Pro is gemini-3-pro-image, listed as stable with no section on the pricing page at all, so there’s no rate to quote for it. Note one more thing. The deprecations page names gemini-3.1-flash-image-preview as the October 2 replacement for gemini-2.5-flash-image, and that ID doesn’t appear on the models page. Pin gemini-3.1-flash-image.
- Image output has a text-and-thinking component as well as a per-image one
- Nano Banana Pro has no published rate: get one in writing before you build on it
- Check text, likeness and rights in every generated asset
Gemini Omni Flash GA claimed, preview listed
The multimodal creation model is gemini-omni-1.1-flash. It adds video extension, interpolation and configurable resolution at 360p, 720p, 1080p and 4k. It’s priced on the paid standard tier at $1.50 per 1M input and $9.00 per 1M output, with no free tier at all. Its status depends which Google page you read. DeepMind and the release notes say GA. The API models page lists it under a preview heading. The August 27, 2026 announcement says only that it’s production-ready and rolling out. The preview endpoint it succeeds, gemini-omni-flash-preview, was due to stop on September 30, 2026. That date has passed with no first-party statement either way, and the pricing page still sells the preview.
- Pin
gemini-omni-1.1-flash, not the preview - Say “Google describes it as production-ready,” not “it is GA”
- If anything you own still calls the preview, test it today rather than trusting the pricing page
Text to speech Stable September 22
Two new voice models went stable, and the preview one became legacy. First, gemini-3.8-flash-tts. It’s Google’s flagship creative text-to-speech model, aimed at studio-grade fidelity, expressive acting, voice design and voice replication. gemini-3.8-flash-lite-tts is the cheap, fast one for high-volume production and real-time voice agent cascades. A voices endpoint and a library Google puts at 150 or more prebuilt and custom voices come with them. Pricing is $0.50 per 1M input and $9.00 per 1M output through December 31, 2026, then $1.00 and $18.00. Voice replication is the feature to govern, not the price. Get written consent from the person whose voice you clone, and keep it.
gemini-3.1-flash-tts-previewis now marked legacy: migrate off it- The output rate doubles on January 1, 2027
- The GA date comes from Google’s release notes, which are rewritten in place
Lyria 3.5 GA September 3
Google’s full-length song generation model, listed on DeepMind’s model index and shipped into the Gemini app. The API models page, read October 1, 2026, lists lyria-3-clip-preview, lyria-3-pro-preview and lyria-realtime-exp, and no Lyria 3.5 ID. Google describes improved musical coherence, natural vocals and fine-grained duration and structural control, generating 44.1 kHz stereo audio from text and image inputs. The September 3 date comes from the API release notes. The announcement page showed no publication date when it was read. So confirm the date directly. Music generation is also listed inside all three consumer AI plans.
- Text and image inputs, 44.1 kHz stereo output
- Rights clearance matters more for music than for most generated media
- Confirm the GA date directly before you cite it
Gemini 3.5 Transcribe GA August 26
Speech-to-text in two forms, gemini-3.5-transcribe and gemini-3.5-transcribe-live, covering 85 or more languages with speaker diarization and custom vocabulary biasing. Vocabulary biasing is the feature that makes internal product names and people’s names survive a transcript. The live model is 75 percent more expensive per 1M tokens than the batch model. Per minute, the gap depends on direction: input is $0.005 against $0.003, which is 67 percent more, and output is $0.004 against $0.002, which is 100 percent more. So the choice between the two IDs has a number attached to it.
- Separate IDs for batch and live, with a real price gap
- Diarization for multi-speaker recordings
- Bias the vocabulary before you judge accuracy
Veo and Flow
Google’s video stack supports cinematic clips and scene-oriented workflows, not just a single generated shot. Google Flow is listed inside the Google AI Pro plan and above.
- Storyboard the intent
- Use references and continuity
- Plan for post-production
Slides and Workspace
Gemini can help move research and narrative into presentation form inside the same organizational environment. The August 26 Workspace Feature Drop adds Meet notes screenshots, automatic meeting-artifact organization in Drive and Forms quiz generation. It was the most recent drop as of October 1, 2026.
- Use brand templates
- Link claims to sources
- Confirm edition eligibility for each drop feature
Gemini Notebook briefing outputs
Audio, video, reports, quizzes and other outputs turn a curated source set into several learning formats. Google reports more than 30 million users and over 600,000 organizations on the product.
- Source-grounded transformation
- Useful for enablement
- Keep audience needs explicit
Google Vids July 16
Vids gained Gemini Omni clip generation, and personal avatars built from a selfie and a voice sample, watermarked with SynthID. Google lists it for Google AI Pro and Ultra, plus Workspace business editions.
- Check plan eligibility first
- Get written consent for any likeness
- Disclose synthetic presenters
Generative UI
Custom charts, tools and simulations are a new kind of creative output for questions that need interaction. Google promised general Search availability “this summer” on May 19. Treat it as promised, not shipped.
- Explain complex systems
- Let users explore variables
- Test the generated behavior
Using the research notebook and brand system, create three campaign territories. For each, provide the strategic idea, hero image direction, six-second video beat and the evidence that makes the concept credible.
Turn this Gemini Notebook source set into an executive presentation. Keep one claim per slide, cite the exact source behind each claim and use our approved Slides theme.
Build an interactive comparison tool for these options using current Search and Maps data. Let the user change budget, distance and priority, and show how the recommendation changes.
Inspect every generated visual for inaccurate text, impossible product details, misleading data or implied claims that the source material does not support.
Ten habits for using Gemini as an ecosystem
Pick the surface that already owns the context. Pin what you build on. Cost the tier the work will actually run on. And keep the evidence attached as the work moves toward action. That’s the whole behavior, right?
Start in the context owner
Use Search, Workspace, Gemini Notebook, Android or the API, depending on where the evidence already lives.
Ask for grounding
Use live Search and Maps when current external reality matters. The free grounding allowance is shared across Gemini 3.x models, so budget it once.
Curate when authority matters
Use Gemini Notebook or a controlled source set instead of the whole web.
Use the actual modality
Give Gemini the image, the video, the voice or the screen state instead of describing it badly. Voice is a first-class production build now, not a demo.
Separate prepare from act
Let Gemini draft first. Require approval before anything consequential happens outside.
Name the surface and the plan
Availability differs across app, Search, Workspace, API, region and subscription, and the consumer plan decides which Pro model and how much of it an account reaches.
Cost the tier, not the headline
Say whether a figure is free tier, paid standard or batch. Batch is exactly half of standard, so a batch workload costed at the standard rate overstates spend by 2x.
Preserve source links
Keep live information attached as it moves into a document, deck or decision.
Discount the superlatives
Google has now called three Flash generations its most intelligent workhorse model. Compare model cards, prices and your own traces instead.
Pin the model, not the alias
Name gemini-3.8-flash or another explicit ID, and check a named replacement against the models page before you adopt it. Aliases move on Google’s schedule, not yours.
An ecosystem this broad is easy to overstate
Three mistakes, in the order people make them. Turning a limited preview into a universal capability statement. Trusting a documentation page to describe the present. And repeating an absence after the vendor has quietly filled it. Gemini features arrive across products, plans and regions at different times, which is what makes all three so easy.
An absence is a claim too
Nobody re-checks a negative. That’s what makes it the dangerous kind. Google’s API model page carries no knowledge-cutoff field for Gemini 3.8 Flash at all, which reads like an unpublished specification. The DeepMind model card published March 2026 on the day the model shipped. A statement that something doesn’t exist expires exactly like a statement that it does. So read the model card and the API reference before you report that a specification is missing.
The vendor can contradict itself
March 2026 on the model card. No cutoff field at all on the API model page. GA on DeepMind and in the release notes for Omni, a preview heading on the models page. A replacement ID on the deprecations page that the models page has never heard of. A model scheduled to stop on October 2 and listed as stable on October 1. A price in a blog post for a model the pricing page doesn’t carry. Name both surfaces and say which one you trusted. A dated article governs the record. A live page governs the present.
Rollout fragmentation
Plan, region, language, hardware and Workspace edition can each change access. And the consumer plan tier now changes which Pro model an account can reach, not just how often.
Prompt counts are the wrong unit
Google moved the consumer plans from daily prompt limits to a compute-used model, where limits factor in prompt complexity, the features used and chat length. Any plan comparison built on prompts per day is measuring something Google no longer sells.
Quote the tier with the number
Every published Gemini rate belongs to a tier. Free tier is free of charge on the Flash text models, and not available on the image models or Omni. Batch is exactly half of paid standard. A batch workload costed at standard overstates spend by 2x.
The cheap model is the one with the end date
Deprecation usually moves you forward and down in price. Not here. gemini-3.1-flash-lite, at $0.25 per 1M input and $1.50 per 1M output, is scheduled to stop serving on May 7, 2027. And Google names the older, dearer gemini-3.5-flash-lite as its replacement at $0.30 and $2.50. So the migration carries a price increase, and the models page still advertises the retiring model as frontier-class performance at a fraction of the cost. Re-cost any high-volume pipeline before you treat the swap as neutral.
“Legacy” is a signal
Google’s models page now calls Gemini 3.5 Flash its legacy Flash model, while 3.6 Flash is merely previous-generation. No shutdown date has been published. But legacy is the word that comes before one, and 3.5 Flash is also the expensive Flash. Plan the move before the notice arrives.
Grounding errors
A cited Search result can still be misunderstood, stale or weaker than the claim.
Computer-use risk
An interface can hand your agent instructions you never wrote. Google now documents prompt injection detection for Gemini 3.x computer use. So there’s a first-party control to turn on, not just a warning to repeat. Two cautions remain. The docs no longer publish a supported-model list, so test model support rather than looking it up. And Agent Anomaly Detection, the runtime oversight layer, was private preview when Google announced it on September 16, 2026.
An agent contract can break under you
Google replaced the default Antigravity agent on September 17, 2026, moving built-in tool parameters to PascalCase and file edits to line ranges. Then it put a date on the old one: antigravity-preview-05-2026 stops serving October 5, 2026. A model swap is visible in a dashboard. A tool-contract change shows up as failing agents, and a shutdown shows up as nothing at all. So read the agent release notes on the same schedule you read the deprecations table.
Background-agent drift
Long-running monitoring needs a clear objective, notification policy and stop condition.
Part of Apple’s private inference now runs in Google Cloud
Apple calls its third-generation foundation models “custom-built in collaboration with Google.” It names five. And for the most capable, AFM 3 Cloud Pro, it says this: “we worked with Google and NVIDIA to extend Private Cloud Compute to NVIDIA GPUs in Google Cloud.” So some of Apple’s private inference runs on Google’s infrastructure, on hardware Apple didn’t build. None of that moves the model and price tables above. But a client may already reach Google through Apple rather than through you. So ask where the model actually runs before you scope a Google agreement.
· Apple Machine Learning Research, June 8, 2026, on the AFM 3 family, and Apple Security Research, “Expanding Private Cloud Compute,” June 8, 2026.
Personal versus work context
Connected apps and device context can cross boundaries employees or administrators didn’t intend.
Generated media truth
Visual quality doesn’t guarantee accurate text, products, scenes or implied evidence. Image and music output also carry rights questions that a text answer doesn’t.
Product-name churn
Build guidance around durable workflows, with dated model notes, rather than temporary launch labels. NotebookLM became Gemini Notebook on July 16, 2026.
The superlative is a constant
Google called 3.6 Flash, then 3.7 Flash, then 3.8 Flash a version of “our most intelligent workhorse model.” A phrase reused across three generations can’t distinguish any of them. Take the differences from the model card and the price table, and never let a vendor adjective carry a recommendation.
Announced is not available
Spark, Gemini 4 Argon, the cyber models, enterprise Live access and generative UI all have to keep their actual rollout state attached. Argon is the newest case: announced with a price on September 30, 2026, and absent from the models page, the pricing page and the model-card library. Spark is the sharpest. Google’s support page, read October 1, 2026, still excludes the European Economic Area, Nigeria, Switzerland and the United Kingdom, and Spark also needs a Pro plan or above.
Earliest possible, not final
The deprecations page says its shutdown dates “indicate the earliest possible dates on which a model might be retired,” with advance notice before actual discontinuation. Plan against the published date, but describe it the way Google does rather than promising a hard deadline Google hasn’t promised.
A shutdown date can be withdrawn, and access can be cut without one
In early August Google’s deprecations page dated the whole Gemini 2.5 family to October 16, 2026, then removed it. Read October 1, 2026, all three models still read “No shutdown date announced.” But Google has since narrowed who can use them: “we are limiting access to the 2.5 models to users who have actively used them in the past.” New projects are pointed at 3.5 Flash-Lite or 3.8 Flash. No shutdown date doesn’t mean a new project can call the model.
A deprecation row can vanish instead of going gray
gemini-omni-flash-preview was dated to stop on September 30, 2026. On October 1 its row is simply gone from the deprecations page, which elsewhere promises that “Already-shutdown models are indicated with gray backgrounds.” The models page no longer lists the endpoint. The pricing page still sells it. No Google page states an outcome. So the deprecations table is a forward schedule, not a history. Keep a dated copy of the row you depend on.
A shutdown date can also pass unconfirmed
gemini-robotics-er-1.6-preview was scheduled to shut down on August 31, 2026. Thirty-one days later the deprecations page still lists it as a future shutdown. The models page still lists it with no deprecation mark. And the robotics overview, edited twice since the deadline and most recently on September 23, still says the model “will be shut down at the end of August.” Two post-deadline edits preserved the future tense. So report the passed date and the disagreement, not an outcome.
A migration target can be on a fuse
The rewritten Imagen page tells users to move to gemini-2.5-flash-image, naming gemini-3.1-flash-image only as a parenthetical alternative. The deprecations page stops the first of those on October 2, 2026. The page hasn’t been edited since September 17, so the advice is still in that order the day before the shutdown. Check every named replacement against the shutdown table, including the ones a fixed page gives you.
Documentation lags its own product
The Imagen page spent weeks describing an August 17 shutdown in the future tense before Google rewrote it on September 17. So use one rule. A dated article is authoritative for what was announced and when. A live page is authoritative for what is true today. When they disagree, the live page governs the present and the article governs the record. Keep your own dated copy of the live page, because it carries no history. And a corrected page is only as current as its correction.
Whose benchmark is it
Google attributes the 17 percent output-token reduction for 3.6 Flash to the Artificial Analysis Index. The 88 percent token, 66 percent cost and 7 percent accuracy figures for agentic video are Google’s own. So are the multipliers advertised on the consumer plans, which are measured against non-AI subscribers. Keep each attribution attached to the party that produced it.
Alias drift
No first-party Google page documents which model gemini-flash-latest serves today. The last documented switch in the release notes remains January 21, 2026, to gemini-3-flash-preview, read October 1, and Google’s stated policy is a hot swap with two weeks of email notice. So pin an explicit model ID. xAI shows both ways an alias bites. On July 29, 2026 it announced that grok-voice-latest would move to grok-voice-think-fast-2.0 on August 5. That reroute was dated and public, and it still raised the per-minute audio rate from $0.05 to $0.08 for anyone on the alias. The unannounced failure came later. xAI took grok-voice-think-fast-1.0 off its docs price table with no deprecation mark and no dated notice.
A missing cadence is information
Two consecutive months have closed with no Gemini Drop. July 31, 2026 remains the most recent edition, and both candidate URL patterns 404 for August and September. That matters because the drops have migrated between paths before. Android and Pixel drops did ship in September, so this gap is specific to Gemini. October isn’t due yet, since the July edition landed on the 31st. Don’t present the drop as a reliable monthly channel, and don’t call it dead until the October window closes.
Slipped timelines, plural
Gemini 3.5 Pro was described on May 19, 2026 as rolling out the next month, and on July 21 as testing with partners. The month Google named was June. It’s October and the model is still absent from the models and pricing pages, and DeepMind’s pages no longer mention it at all. Gemini 3.1 Pro shipped in preview on February 19, 2026 with general availability promised soon, and it’s still a preview endpoint seven and a half months later. Two slipped Pro timelines is a pattern, not an accident, and one of them has now gone quiet rather than late.
A user-facing feature can be deprecated too
Deprecation isn’t only an API problem. Google replaced Gems with Skills on September 30, 2026 and dated the retirement: November 2026 for personal accounts, March 2027 for Workspace business, enterprise and nonprofit, June 2027 for Workspace education. Anyone in your organization who built a Gem has built it on something with an end date. Inventory them before November.
| Claim type | Before publishing | Safe wording |
|---|---|---|
| Model default | Check the official model release page | “Default in the Gemini app as of…” |
| Model identifier | Check the release notes for the last documented alias change, and check a named replacement against the models page | Name the pinned model ID and its GA date, never an alias |
| A published absence | Re-check the model card before repeating that Google has not published something | Say what the card publishes and on what date you read it |
| Price | Identify the tier: free, paid standard or batch, and the context tier for Pro | Name the tier in the same sentence as the number |
| Consumer plan | Find the dated post that states the amount; the plans page renders prices client-side | Quote the figure with the date Google stated it, and keep the “price varies by country” caveat attached |
| Vendor superlative | Check whether the same phrase covered the previous model | Quote it with the date and the model it was written about |
| New agent | Check plan, region, tester status and the agent tool contract | “Rolling out to…” or “available to…” |
| Workspace feature | Check edition and admin controls | Name the Workspace surface and eligibility |
| Computer use | Test the model ID you intend to pin; the documentation no longer publishes a supported-model list | Name the ID you verified and the date you verified it, and say that safety controls including prompt injection detection are configurable |
| Creative capability | Check model, product and output availability, and the per-image rate | Separate announced model from shipping user feature |
| Model lifespan | Re-read the deprecations page rather than trusting a date you recorded earlier | Give the shutdown date as the earliest possible date, the named replacement and the date you read the page |
| An executed shutdown | Look for a page that says it happened, not one that says it will | Say the date passed and name the pages that still disagree |
| Regional availability | Read the exclusion list, not the launch framing | Name the excluded regions in the same sentence as the launch |
Where these facts come from
All but two of these Google announcements carry a full publication date. The Gemini 3.7 Flash model card carries no date at all. The Gemini 3.5 Transcribe announcement is stamped August 2026 with no day. Treat those two as undated first-party pages rather than dated citations. Google’s model lineup moves across the app, Search and the API at different speeds, so a claim that’s true of one surface isn’t automatically true of another. And a large part of what a buyer needs from Gemini lives on pages Google rewrites in place. One of those is listed below, with the date it was read, because it’s the only record Google publishes of when a model dies. The rest are attributed inline, and anyone relying on one should confirm it directly. Here is that list. The paid-tier standard, free-tier and batch prices, and every figure in the price table, from the pricing page, which carries no last-updated stamp at all. The models-page descriptions, including Gemini 3.5 Flash as “legacy,” the stable listings for gemini-3.1-flash-lite, gemini-3.1-flash-image, gemini-3-pro-image, gemini-embedding-001, lyria-3-clip-preview, lyria-3-pro-preview and lyria-realtime-exp, and the preview listing for gemini-omni-1.1-flash. Lyria 3.5 appears on the DeepMind model index, not on the models page. The absence of any Gemini 4 Argon entry from the models page, the pricing page and the DeepMind model-card library. The Imagen page and the robotics overview. The computer-use page, including the three environments and the prompt injection detection, which carries no stamp at all. The Spark availability sentence on Google’s support page. The consumer plan contents on Google’s AI plans page, whose prices render client-side. And five items exist only in the API release notes. The September 22, 2026 general availability of the two Gemini 3.8 TTS models. The September 17, 2026 replacement of the Antigravity agent by antigravity-preview-09-2026. The September 3 general availability of Lyria 3.5. The August general availability dates for gemini-omni-1.1-flash and the Gemini 3.5 Transcribe models. And the January 21, 2026 switch of gemini-flash-latest to gemini-3-flash-preview. One more. The Lyria 3.5 announcement page is named inline rather than listed. It didn’t surface a publication date when it was read, so it can’t serve as a dated citation.
AI Mindset