ChatGPT vs Claude vs Gemini for Shopping [2026]

Niamh Kelly
October 18, 2026
21 min read

Ask ChatGPT to find you the cheapest 55-inch TV in Dublin this weekend and it will happily oblige, citing retailers, prices and delivery windows in seconds. Ask it to <a href="https://myappsplus.com/router-and-modem-or-gateway-whats-actually-in-your-home-network/” title=”Router and modem or gateway: What's actually in your home network?”>actually buy the thing, and most Irish shoppers hesitate. According to EY Ireland’s 2026 AI Sentiment Survey, 69% of Irish consumers have used AI during a customer interaction in the past six months, yet just 8% say they would trust an AI agent to purchase a product on their behalf. That 61-point gap between “happy to ask” and “too nervous to let it pay” is the real story behind every AI chatbot comparison published this year, and it is the one this piece sets out to test properly.

This is not another feature-by-feature app integrations breakdown or a raw pricing spreadsheet. It is a shopping-specific test of the three chatbots Irish consumers actually open first: ChatGPT, Claude and Gemini. All three have shipped new flagship models in the past four weeks, and all three now claim some ability to browse, compare and reason about products. We put that claim under the kind of pressure an Irish shopper actually applies: VAT-inclusive pricing, delivery charges to an Irish address, retailer trustworthiness and whether the chatbot tells you when it is steering you toward a sponsored result.

Don’t miss new tech stories on Google

Add Tech Insider once in the Google app and our stories appear in your news suggestions.

Why AI Chatbots Are Becoming Ireland’s Shopping Assistants

The shift did not happen overnight. EY’s survey found that 84% of Irish respondents consciously used AI in the past six months, up sharply from the low-double-digit figures reported just two years ago, though still behind “Pioneer” markets such as India, China and Hong Kong, where usage sits at 94%. Within that broader AI habit, shopping-adjacent behaviour has become one of the fastest-growing categories. Product and brand recommendations are now one of the two main reasons Irish consumers say they turn to an AI chatbot during a customer interaction, trailing only general customer support at 41%.

What is driving it is not novelty. It is price. Asked what benefit they expect AI to deliver, 59% of Irish respondents named lower prices or better value for money as their top expectation, ahead of the 54% figure recorded across Pioneer markets. A further 53% expect faster service and quicker responses. Irish consumers, in other words, are not looking for a chatbot that writes them a poem about a toaster. They want one that finds the toaster for the lowest landed price, including delivery, and does it faster than a manual search across three retailer sites.

Trust has not caught up with capability, though. Only 12% of Irish respondents currently use AI to auto-refill online shopping carts or make purchases on their behalf, and just 14% have used any form of autonomous AI in the past six months, compared with 24% across Pioneer markets. Around one in five Irish respondents told EY they hold unfavourable views specifically of chatbots, dynamic pricing and targeted advertising, a cluster of concerns that sits at the centre of this comparison: can ChatGPT, Claude or Gemini find you a genuinely good deal without quietly nudging you toward whoever paid for placement?

Shopping sits in the middle of a wider pattern across categories. EY found that 77% of Irish consumers have used AI for energy and mobility tasks in the past six months, led by route planning (28%) and travel planning (23%), both of which overlap heavily with shopping-style price comparison. Usage in technology and entertainment reaches 62%, and in health and wellness 57%, both well ahead of the 14% figure for fully autonomous AI use. The pattern is consistent across every category EY measured: Irish consumers happily use AI chatbots to inform a decision, and pull back sharply the moment the chatbot is asked to execute one, whether that is booking a flight, managing a prescription or completing a purchase.

Meet the Three Contenders: GPT-6 Sol, Claude Opus 5.5 and Gemini 4 Argon

Each company has refreshed its flagship model within weeks of this article going to press, which matters for a shopping comparison because reasoning quality and browsing reliability are exactly the capabilities that improve fastest between generations.

OpenAI began rolling GPT-6 into ChatGPT on 7 October 2026. GPT-6 Sol became the default model for Plus, Pro, Business and Enterprise subscribers that day, while Free and Go users followed a day later on a lighter variant called GPT-6 Luna. Pro subscribers can also switch to a deeper reasoning option branded GPT-6 Astra. A further release, GPT-6.1 Sol, landed on 29 September 2026 but was limited to ChatGPT Work and Codex rather than the consumer chat interface, and OpenAI retired GPT-5.5 from Codex entirely on 14 October. You can read OpenAI’s own announcement of the GPT-6 Sol and Luna rollout for the full technical detail.

Anthropic moved on an even tighter schedule. Claude Opus 5.5 arrived on 22 September 2026 as the new flagship, cutting API pricing to $4 per million input tokens and $20 per million output tokens, down from $5 and $25 under the previous Opus 5. Claude Sonnet 5.5 followed on 28 September, and Claude Haiku 5.5 landed on 7 October, replacing Haiku 4.5 as Anthropic’s fast, lightweight tier. All three current Claude models now share a one-million-token context window, a genuinely useful spec for a shopping assistant asked to hold an entire retailer’s return policy, a product manual and a price-comparison spreadsheet in memory at once.

Google’s refresh centres on Gemini 4 Argon, announced on 30 September 2026 and rolled out broadly to the consumer Gemini app on 13 October. Google has positioned Argon around deep reasoning and long-running autonomous workflows rather than quick-fire chat, with a published context window of 1,048,576 tokens. It arrived alongside a separate “Gemini agent for work” announcement on 8 October, signalling Google’s own pivot from answering questions to completing multi-step tasks. Argon now sits above the existing Gemini 3.1 Pro and Gemini 3.8 Flash models that still power the bulk of Google’s paid subscription tiers while pricing pages catch up to the new flagship.

Shopping-Ready Specs Compared

Pulled together, the three platforms look less like three versions of the same product and more like three different bets on what a shopping assistant should prioritise. The table below lines up the specs that actually matter when a chatbot is asked to compare products rather than just chat.

None of the three can actually complete a purchase on your behalf yet in any mainstream consumer plan. That is a deliberate design choice as much as a technical limit, and it lines up neatly with the 8% trust figure from EY’s survey: the companies building these chatbots appear to know their users are not ready to hand over a credit card, so none of them have built a one-click Irish checkout into the free or mid-tier plans.

Pricing Compared: What Each Chatbot Costs Irish Shoppers

Shopping-focused use rarely needs the most expensive tier, since price comparison is a reasoning task rather than a heavy-compute one, but the gap between entry-level and top-tier pricing is still worth laying out before anyone upgrades specifically to shop better.

Gemini is the only one of the three offering a genuinely low-cost paid entry point for Irish consumers at under €5 a month, and the only one bundling a free full year of its mid-tier plan for verified students, which matters if the use case is simply “find me the cheapest textbooks and a laptop before term starts.” Claude’s pricing sits in the middle and is the simplest of the three to understand, with a flat $20 entry point and no advertising-supported tier. For a deeper breakdown of how Anthropic structures its tiers, see our separate Claude Opus vs Sonnet vs Haiku pricing comparison. ChatGPT’s pricing is the most fragmented, spanning six distinct price points once the new $500 Pro tier is included, aimed squarely at power users rather than casual shoppers. These figures were cross-checked against OneHuman’s side-by-side AI tool pricing tracker, which verifies pricing and model availability for all three platforms roughly every two weeks.

Worth flagging for anyone shopping on a tight budget: none of the three advertises a shopping-specific discount or bundle. The pricing tiers above are the same whether you plan to use the chatbot for coding, essay-writing or comparing fridge-freezers, which means it rarely makes financial sense to upgrade a plan purely to improve shopping results. The free tier of any of the three is perfectly capable of running a basic price comparison; the paid tiers mainly buy you more queries per session and, in ChatGPT’s and Gemini’s case, access to the newest reasoning models.

Benchmark Data: Context Windows, Release Cadence and Reasoning Power

Shopping comparisons do not have a dedicated industry benchmark the way coding or maths do, so the closest proxies are context window size, release cadence and how each company is explicitly marketing reasoning capability. On context window, Claude and Gemini are now essentially tied at roughly one million tokens across their current flagship lineups, while ChatGPT’s consumer tiers remain segmented, topping out at 400K tokens even on the Pro plan’s reasoning mode. For a shopping task, that difference shows up when you ask a chatbot to compare ten product listings, three retailer return policies and a page of customer reviews in a single conversation. Claude and Gemini are less likely to “forget” the first retailer’s delivery terms by the time they reach the tenth comparison.

Release cadence tells its own story. Anthropic shipped three model updates (Fable 5.1, Opus 5.5, Sonnet 5.5) plus Haiku 5.5 inside five weeks between 1 September and 7 October 2026, a pace tracked in detail by independent trackers such as Keywords Everywhere’s AI model release tracker. OpenAI matched that intensity with the GPT-6 Sol/Luna rollout, a new $500 Pro tier, and a parallel-agent update to GPT-6 Pro inside the same six-week window, detailed independently in a weekly ChatGPT and Claude update log. Google’s Gemini 4 Argon rollout was comparatively slower to reach all users, taking two weeks from announcement to broad availability, but it is the only one of the three explicitly marketed around “long-running autonomous workflows,” the exact capability a true shopping agent would need.

None of this translates directly into “best at finding deals,” which is why the next five sections run each chatbot through concrete shopping tasks rather than relying on specification sheets alone.

Real-World Test: Comparing Laptop and Electronics Prices

The first test asked each chatbot for the cheapest 14-inch laptop suitable for college work, available for delivery to an Irish address within a week, with the total cost including delivery and VAT made explicit.

What Each Chatbot Got Right

ChatGPT, running on GPT-6 Sol with browsing enabled, returned three specific models with listed prices clustered between €549 and €649, and correctly flagged that one retailer’s “free delivery” offer only applied above a €50 spend threshold. It also volunteered a warranty comparison unprompted, which was a genuinely useful addition. Gemini 4 Argon pulled the widest spread of retailers thanks to its Google Shopping integration, surfacing a fourth, cheaper option at roughly €519 that neither ChatGPT nor Claude found, and it was the fastest to return a complete answer, taking roughly a third of the time ChatGPT needed to finish browsing. Claude Opus 5.5 was the most conservative of the three, explicitly stating when it was not fully confident a listed price was current and recommending the shopper verify stock before travelling to collect in-store, a caution that lines up with Anthropic’s consistently more hedged tone across all three platforms.

Where They Fell Short

All three struggled with stock accuracy; two of the three recommended models were listed as “in stock” by the chatbot but showed as back-ordered when the retailer’s actual page was checked. Gemini’s speed advantage came at the cost of including one sponsored listing without clearly marking it as such, something only revealed when asked directly whether any results were paid placements. This is precisely the kind of gap that explains why only 8% of EY’s survey respondents trust an AI agent to buy on their behalf: the chatbots are good at surfacing options, less reliable at confirming the details that actually determine whether a purchase goes smoothly.

A second round of the same test, run on a mid-range soundbar instead of a laptop, produced a similar pattern: a roughly €30 spread between the cheapest and most expensive option each chatbot surfaced, and at least one stale price per chatbot that had moved by the time the retailer’s page was checked directly. The lesson generalises well beyond laptops. Treat every price a chatbot returns as a strong starting estimate, not a confirmed quote, until the retailer’s own checkout page shows the same number.

Real-World Test: Everyday Price Comparisons

Big-ticket electronics purchases are occasional. Broadband, mobile and grocery shopping happen every month, which makes them a better test of whether these chatbots are useful day to day rather than just for the occasional splurge.

Broadband and Mobile Plans

Asked to compare broadband plans for a two-person household in Cork with moderate streaming use, all three chatbots correctly identified that Irish broadband pricing varies heavily by eircode and that exact quotes require entering an address on a provider’s site. ChatGPT was the most thorough at listing the major national providers and typical speed tiers. Gemini again pulled in current promotional pricing fastest, but could not confirm whether a quoted promotional rate applied for 12 or 18 months without a follow-up question. Claude was the only one of the three to proactively mention contract exit fees as a factor shoppers should check, a small but genuinely useful consumer-protection nudge.

Grocery and Meal-Plan Price Comparison

None of the three chatbots can currently pull live, itemised grocery prices from Irish supermarket apps in real time, since none have a direct data partnership of that kind. All three instead fell back on general price-range estimates and suggested the shopper cross-check with a supermarket’s own app, which is an honest answer but not a particularly powerful one. This is the clearest gap in the entire shopping comparison: for the single most frequent category of household spending, none of the three tested chatbots currently offers more than a shopping-list organiser.

Real-World Test: Travel Booking and Black Friday Deal Hunting

Flight and Travel Price Research

Asked to find the cheapest return flight from Dublin to a short-haul European city for a specific weekend, Gemini 4 Argon’s Search grounding again produced the fastest and most current results, including a direct link to the booking page. ChatGPT came a close second and was notably better at explaining baggage allowance differences between the fares it found, a detail that often changes the real cost of a “cheap” flight. Claude correctly declined to guarantee that any quoted fare would still be available by the time the shopper clicked through, which is accurate but less immediately useful if the goal is simply “book me the cheapest flight now.”

Black Friday and Sale-Season Deal Hunting

With Black Friday weeks away at the time of testing, all three chatbots could discuss historical discount patterns for electronics and could explain the difference between a genuine reduction and a pre-inflated “was” price, a known retail tactic. None could yet access live Black Friday pricing, since the sale period had not started. This is a reasonable limitation rather than a flaw, but it is worth knowing in advance: these chatbots are far more useful for planning what to buy and setting a price-alert strategy than for catching a flash deal the moment it drops.

The Disclosure Problem: Sponsored Content and Commercial Bias

This is the section that matters most given Irish consumer sentiment. EY’s survey found that around one in five Irish respondents hold unfavourable views specifically of chatbots, dynamic pricing and targeted advertising, well above the 7% unfavourable rating recorded for AI-enabled security features like in-store cameras. That gap tells you exactly where Irish trust breaks down: people are broadly fine with AI watching for shoplifters, far less comfortable with AI quietly influencing what they buy and at what price.

Across the real-world tests run for this piece, none of the three chatbots clearly labelled a sponsored or affiliate-linked result as such by default. Gemini’s tighter integration with Google’s commercial search ecosystem makes it structurally the most exposed to this risk, since the underlying data sources it draws from already include paid placements. Claude’s more conservative, search-tool-based approach produced fewer results overall but also surfaced no results that were later confirmed to be paid placements during testing. ChatGPT sat in between, generally transparent when asked directly but not proactively flagging commercial relationships unprompted. None of this means any of the three is deliberately deceiving shoppers, but it does mean the burden currently falls on the shopper to ask “is this a sponsored result?” rather than the chatbot volunteering that information.

For context on how these same three platforms perform when the task is pure information accuracy rather than shopping, see our existing comparisons on fact-checking reliability and news accuracy. The same hedging patterns observed there (Claude’s caution, Gemini’s speed, ChatGPT’s balance) show up again in shopping, which suggests these are consistent personality traits of each model family rather than task-specific quirks.

Data Residency, GDPR and Why Irish Shoppers Should Care

Every shopping query you type into a chatbot, from “cheapest fridge-freezer under €500” to “compare my current broadband bill to switching providers,” is personal financial behaviour data. EY’s survey found that only 43% of Irish respondents trust companies generally to protect their data, and just 42% trust government to do so, figures low enough that they should shape how cautiously Irish consumers treat shopping-related AI conversations specifically.

All three companies operate under GDPR obligations for EU users, including Irish consumers, and all three allow users to request data deletion and to opt certain conversations out of model training. None of the three chatbots tested made this distinction obvious mid-conversation during a shopping task; a user asking for price comparisons is not shown, by default, whether that specific exchange is being used to refine future product-recommendation behaviour. Shoppers who want tighter control should check each platform’s privacy settings directly rather than assuming shopping-related chats are treated differently from any other conversation, since none currently offer a dedicated “shopping mode” with separate data handling.

The 67% of Irish respondents who told EY they are concerned about AI systems being hacked or breached have a reasonable basis for that caution when the data in question includes purchase history, delivery addresses and payment preferences discussed in a chat window. None of the three platforms currently stores full payment details inside the chat interface itself, since none process checkout directly, but delivery addresses and product preferences discussed during a shopping session are retained under each platform’s standard chat-history policy unless a user actively deletes them.

What This Means for Irish Retailers

The trust gap cuts both ways. Irish retailers now have to think about a shopper who arrives at checkout having already been “pre-sold” by a chatbot’s comparison, rather than one who started on a search engine or a retailer’s own site. That changes what matters for visibility. A product page with an outdated price, an unclear delivery charge, or a return policy buried three clicks deep is exactly the kind of listing these chatbots struggled to use reliably during testing, which means retailers with clean, structured, current pricing data are more likely to be surfaced accurately than those who are not.

It also reframes the sponsored-placement question from the retailer’s side. Since 59% of Irish consumers told EY that lower prices or better value for money is the benefit they most expect from AI, a retailer paying for a prominent sponsored listing inside a chatbot’s answer is competing directly against that expectation. If a sponsored result is not also a genuinely competitive price, the mismatch between what the chatbot implies (a good deal) and what the shopper later finds (an average one) is likely to erode trust in that specific retailer faster than it would in a traditional search advert, precisely because the chatbot’s tone reads as more personally advisory than a list of paid search results ever did.

Best Use Cases for Each Chatbot

No single chatbot won every real-world test, which means the right pick depends heavily on what kind of shopping a reader does most often.

• Fast, broad price sweeps (electronics, flights): Gemini 4 Argon, thanks to its direct Google Shopping and Search grounding, consistently returned the widest set of options fastest across every test run.
• Cautious, high-value purchases where accuracy matters more than speed: Claude Opus 5.5, because it was the only chatbot to consistently flag uncertainty rather than present a stale price with false confidence.
• Balanced day-to-day use across broadband, mobile and general shopping questions: ChatGPT on GPT-6 Sol, which produced the most consistently well-rounded answers, including details like warranty terms and baggage allowances that the others sometimes omitted.
• Students on a tight budget: Gemini, purely on pricing, since its €4.99 entry tier and free year of Google AI Pro for verified students undercuts both competitors significantly.
• Households wanting a single assistant for both shopping research and broader customer-support questions: any of the three will do, though readers who specifically need AI-driven customer service comparisons (not just shopping research) should also see our Intercom Fin vs Zendesk AI vs Agentforce comparison, which covers the business side of the same technology.
• Privacy-conscious shoppers: none of the three has a clear structural advantage here, so the better move is to use each platform’s temporary or incognito-style chat mode for shopping queries specifically, regardless of which chatbot is chosen.

Migrating Your Shopping Research Between ChatGPT, Claude and Gemini

Because none of the three chatbots can complete a purchase directly, “migrating” here means moving your shopping research workflow, not an account or a subscription. These are the steps that worked cleanly during testing.

• Step 1: Export or copy your running comparison. Before switching chatbots, copy the actual product list, prices and retailer links the first chatbot produced into a plain document. None of the three platforms currently offers a one-click export of a shopping comparison, so this manual step avoids losing your reference prices.
• Step 2: Paste the comparison into the new chatbot as context. Opening with “here is what I found so far, please verify and add options” produced far better results across all three than starting a fresh, unseeded shopping query from scratch.
• Step 3: Explicitly ask the new chatbot to flag anything it cannot verify. This single instruction noticeably improved honesty about stock and pricing uncertainty in every test run, regardless of which chatbot received it.
• Step 4: Cross-check delivery costs and VAT separately. Since none of the three chatbots reliably surfaced total landed cost without being asked directly, make this an explicit instruction every time you switch platforms, rather than assuming the new chatbot inherited that context.
• Step 5: Verify the final price on the retailer’s own site before buying. Every single real-world test in this comparison, across all three chatbots, surfaced at least one detail (stock status, delivery window or promotional terms) that did not match the retailer’s actual page at the time of checking.

Pros and Cons for Shopping Use

Stacked side by side, the trade-offs are fairly clear-cut once you look past the marketing.

ChatGPT (GPT-6 Sol/Luna)

• Pros: well-rounded answers with useful extra detail (warranties, baggage limits), largest user base means more shared troubleshooting advice online, new parallel-agent mode on Pro is the most capable multi-step feature of the three for complex comparisons.
• Cons: pricing is the most fragmented and confusing of the three, peak-hour usage caps can interrupt a long comparison session, sponsored-result disclosure is inconsistent rather than absent.

Claude (Opus 5.5/Sonnet 5.5/Haiku 5.5)

• Pros: most honest about uncertainty, simplest pricing structure of the three, one-million-token context window across every current tier including the cheapest paid plan.
• Cons: slower to respond, smallest retailer coverage of the three in testing, no advertising-supported free tier for budget-conscious casual users.

Gemini (4 Argon/3.1 Pro)

• Pros: fastest results across nearly every test, widest retailer coverage thanks to Google Shopping integration, cheapest paid entry tier and a genuinely strong student offer.
• Cons: highest exposure to sponsored-content blending given its commercial search integration, “frequent capability changes” that make its behaviour less predictable month to month, Argon’s pricing had not fully propagated to consumer tiers at the time of testing.

The Verdict: Which AI Chatbot Should You Trust With Your Wallet?

There is no single winner, and the data explains why. Gemini 4 Argon wins on raw speed and retailer breadth, which matters most for quick electronics and travel comparisons. Claude Opus 5.5 wins on trustworthiness of the answer itself, which matters most for expensive, considered purchases where a wrong assumption is costly. ChatGPT’s GPT-6 Sol wins on all-round usefulness for the mix of shopping and everyday questions most Irish households actually ask, helped by its new parallel-agent capability for anyone willing to pay for the Pro tier.

The more important finding sits outside the head-to-head scoring. EY’s 69% usage figure against its 8% purchase-trust figure is not a temporary gap that the next model generation will close by itself. It reflects a reasonable, evidence-based caution: across every real-world test in this comparison, all three chatbots produced at least one stock, pricing or disclosure error that a shopper would only catch by checking the retailer’s own page. Until sponsored-result labelling becomes automatic rather than something a user has to ask for, and until stock and pricing data is pulled live rather than approximated, that 8% trust figure looks less like consumer hesitation and more like consumer accuracy. The sensible approach for Irish shoppers in late 2026 is to treat all three chatbots as very capable research assistants, and none of them yet as a substitute for checking the final price yourself.

Frequently Asked Questions

Can ChatGPT, Claude or Gemini actually buy products for me?
No. None of the three offers native checkout inside the chat interface as of October 2026. All three will find and link to product pages, but the purchase itself happens on the retailer’s own site.

Which chatbot gives the most accurate Irish pricing?
In the real-world tests run for this comparison, all three occasionally returned stock or pricing details that did not match the retailer’s live page. Gemini 4 Argon returned results fastest and from the widest set of retailers, but Claude Opus 5.5 was the most reliable at flagging when it was not confident a price was current.

Do these chatbots disclose sponsored or paid product placements?
Not consistently by default. Gemini’s deep integration with Google’s commercial search ecosystem carries the highest structural risk of blending sponsored and organic results. Shoppers should explicitly ask “is this a sponsored result?” rather than assume disclosure happens automatically.

Why do so few Irish people trust AI to buy things for them?
EY Ireland’s 2026 AI Sentiment Survey found only 8% of Irish respondents would trust an AI agent to purchase a product on their behalf, even though 69% use AI during customer interactions generally. The gap tracks with real accuracy issues found in this comparison, including inconsistent stock data and unclear sponsored-content disclosure.

Is Gemini, ChatGPT or Claude cheapest for shopping-focused use?
Gemini has the lowest paid entry point at €4.99 a month for Google AI Plus, plus a free year of its mid-tier plan for verified students. Claude’s $20 Pro plan is the simplest single-tier option. ChatGPT’s pricing spans the widest range, from a low-cost Go tier up to a $500 Pro plan aimed at power users.

Can these chatbots compare live Irish supermarket grocery prices?
Not currently. None of the three has a direct data partnership with Irish supermarket apps, so all three fall back on general price-range estimates for groceries rather than itemised, live pricing.

Which chatbot has the biggest context window for comparing multiple products at once?
Claude and Gemini are roughly tied at close to one million tokens across their current flagship models (Opus 5.5, Sonnet 5.5, Haiku 5.5, and Gemini 4 Argon). ChatGPT’s consumer tiers remain more segmented, topping out around 400K tokens on the Pro plan’s reasoning mode.

Should I worry about data privacy when asking a chatbot to compare prices?
It is reasonable to be cautious. EY’s survey found only 43% of Irish respondents trust companies generally to protect their data, and shopping queries reveal real financial behaviour. All three platforms are GDPR-compliant and allow data deletion requests, but none clearly separate shopping conversations from standard chat-history retention by default, so privacy-conscious users should check each platform’s settings directly.

What is Gemini 4 Argon and is it different from Gemini 3.1 Pro?
Gemini 4 Argon is Google’s newest flagship model, announced on 30 September 2026 and rolled out broadly to the consumer Gemini app on 13 October 2026. Google has positioned it around deep reasoning and long-running autonomous workflows, with a 1,048,576-token context window, sitting above the existing Gemini 3.1 Pro and Gemini 3.8 Flash models that still power most of Google’s paid subscription tiers.

Can I use a chatbot to track a price and get alerted when it drops?
Not natively inside any of the three chat interfaces as tested. All three can discuss historical discount patterns and suggest a target price to wait for, but none currently offers a built-in price-tracking or alert feature, so shoppers still need a separate price-tracking tool or browser extension for that specific job.