BNPL, Qomodo, fintech

Here's what to know to raise a Series A right now

by with No Comment Tech

BNPL, Qomodo, fintech

Image Credits: twomeows

There is good news and just “OK” news.

The good news is that the venture capital market is showing signs of stabilizing. The bad news is that raising a Series A will continue to be difficult for founders, especially as venture firms face liquidity problems, higher interest rates, and pressure from their limited partners to be more cautious in their dealmaking.

In 2020, TechCrunch+ reported that founders should start fundraising when they have at least six months of runway left and that they should budget fundraising to last at least three months, with a one-month prep time to a two- to six-week pitch process with investors.

Today, Jesse Randall, the founder of the platform Sweater Ventures, said founders should start looking to raise a Series A when they have about 12 to 15 months of cash runway left.

“Don’t wait any longer than that,” he told TechCrunch+. “The fundraising cycle, once you start it, takes twice as long and requires three times the conversations.”

Leslie Feinzaig, founder of Graham & Walker, says she primarily invests in pre-seed and seed rounds but tells her founders they should start focusing on their business at least 12 to 18 months before fundraising a Series A. This includes understanding their business model, connecting with the proper investors, and stress testing their readiness. The advice investors gave for a Series A this year shows how little and how much everything has changed in the market: Metrics will always be important, but starting early for this longer journey is key.

“In this market, you have to prep for an A way in advance,” Feinzaig told TechCrunch+, adding that it could be fruitful to do so right after closing a seed round. “Time goes by fast, and in my experience, this catches a lot of founders unaware. Focus on your metrics immediately.”

It’s an investor market out there

This year is set to be much different than last year, Randall said.

He’s expecting the market to swing more in favor of investors, and data is showing that VCs are already back in control: The average dollar amount and valuation accompanying a Series A round has dropped, and it’s taking much longer for founders to raise them. Arjun Kapur, a managing partner at Forecast Labs, said that right now, investors are really looking for a viable path toward profitability, an early indicator of which is strong customer retention.

They want companies that do not “have to rely solely on acquiring new customers from marketing with the promise of improved retention in the future,” he told TechCrunch+. “The bar is raised for all founders, and [product-market fit] continues to be what the market looks to see.”

For example, his firm likes businesses that can reach operating expense leverage and have healthy unit economics. For SaaS or subscription-based companies, the firm is looking for a customer retention rate after 13 months, meaning more than the length of a one-year contract. “It’s a view of how sticky the product is,” he continued.

SaaS remains the most popular type of company for investors, according to the latest Carta report. The data shows that, outside of artificial intelligence, the industries that received the most investor attention last year were SaaS, biotech, healthcare, energy and hardware. In 2022, fintech and consumer were in the top spots but have since entered a slump.

Karlos Bledsoe, a principal at Seae Ventures, said to be cautious of trying to time the markets with your fundraising timelines and that it’s best to raise what the company realistically needs to survive in the long-term.

However, the difficulty in raising in some markets could leave some founders wary, especially people who are nervous about down rounds. Kapur suggested that those who previously raised at a high valuation and who are scared to take on a down round should look to find alternative forms of capital to give their businesses more time to grow into their past valuation.

Companies that are not yet ready for Series A should look at raising a small extension round to help them reach profitability or buy more time. “If your company can keep going indefinitely without additional venture capital, you’ll be in a much stronger position to raise,” Feinzaig said.

Pitch decks

Founders hoping to raise a Series A should also pay close attention to their pitch decks, investors said.

Bledsoe said his firm likes to know who is on a team and how each individual strengthens the company. They also ask why a founder is pitching and how the funds will help build out the next round. Understanding the total addressable market, margin potential, and a business’s understanding of its five-force elements — which include knowing the buying power of a product and competition — are also critical.

“You should have a clear direction for your company with early but strong customer points,” he said. “Pitch the big vision you had at the company’s inception, but add data points and refinement from your initial commercial experiences.”

Pitch Deck Teardown: The perfect pitch deck

Feinzaig tells her founders that a robust spreadsheet can be more impactful than a good-looking deck at Series A. “You want metrics that tell the story of how much a customer is worth and how much it costs you to find them, convert them and serve them,” she said, adding that topline numbers can open a door but won’t get a founder far at this point. “Ideally, you have high-value customers, a go-to-market motion that is repeatable and cost-effective to get more of those high-value customers, and with investments in tech, your margins meaningfully improve over time.”

Decks at this point should have data and metrics from the operations side of the business, but nothing else really matters, Randall said. It’s important to note that in this market, the days of founders raising off potential are long gone. “Business outcomes have to do the talking,” he said. “Not the pitch deck.”

An uneven but optimistic market

General advice on raising is usually more nuanced for women and Black founders, as they face more bias from investors. Last year, funding in the U.S. to all-women teams dropped from 2% to 1.8%, while funding to U.S. Black founders dipped from around 1% to 0.48%. The numbers show that the market downturn has had a deeper impact on these two already marginalized groups, meaning the stakes will be higher for fundraising this year.

“While the benchmarks have moved for founders at every level, the unconscious bias that plagues the industry certainly shifted more against underrepresented groups,” Randall said. When these groups pitch, their presentation needs to be together, clearer and more compelling than their counterparts, he said. “This is by no means a reason for these groups not to go out and fundraise, but rather acknowledge that it’s harder than ever and commit to power[ing] through it.”

Often in times of a downturn, investors tend to retreat back to the comfort of their networks, Bledsoe said. This could have a detrimental impact on the Black founders and women who lack these networks but can also really help those who do. “We have seen some benefit from women and Black founders leaning on their networks to serve as references or make warm intros to new investors rather than, or at least in addition to, cold outreach,” he said.

Overall, though, investors are excited about the early-stage markets, even as uncertainty continues to loom. Feinzaig said this year would see “amazing vintage,” while Kapur and Bledsoe said they remained bullish on early-stage innovation and investments in the space.

Randall, who was once a founder himself, expects founders to continue feeling a shock as the market shifts in favor of investors. After all, the year just started, and more changes are bound to happen. “I expect a lot of founders to fight this shift,” he said. “Just don’t fight so hard that you paint yourself into a corner of desperation.”

A seed sprouts from the soil.

Seed funding: Everything founders should know about fundraising, seed rounds and more for 2024

by with No Comment Tech

A seed sprouts from the soil.

Image Credits: wuviveka / Getty Images

If you’re looking at the current seed funding climate and thinking it’s rough out there, you’re not alone. The last few years have been a roller coaster for startups. First came the uncertainty in the early days of the pandemic, then came the exuberance mid to late in the pandemic when cash flowed freely to startups of nearly every stripe. Seed funding sizes were up, and so were valuations.

Today, things aren’t quite so copacetic. Money is tighter, and the hurdles for startups are higher. But for entrepreneurs early in their journey, that doesn’t mean it’s not a good time to raise a seed round.

“I’ve been really excited by the types of entrepreneurs that we’ve been meeting in the seed stage ecosystem right now,” Talia Goldberg, partner at Bessemer Venture Partners, told TechCrunch+. “In some ways, when the markets are down a bit, the real entrepreneurs come out.”

To understand what’s happening with seed rounds this year, TechCrunch+ spoke with Goldberg and two other seasoned investors: Pae Wu, general partner at SOSV, and Maren Bannon, partner at January Ventures. They offered their perspectives on what milestones they look for when evaluating seed-stage pitches, what sorts of round sizes and valuations they’re seeing, and what advice they’re giving their portfolio companies.

Seed round: current mood

The definition of a seed-stage startup has been evolving over the years as round sizes and valuations creep higher. Investors are also expecting to see a bit more from prospective companies, in terms of market fit and revenue. The pandemic is partly to blame, Bannon told TechCrunch+.

“There was a lot of capital in the COVID era that came in — all these angel funds, operator funds, rolling funds, a lot of that was spreading capital at pre-seed,” she said.

As a result, pre-seed valuations were higher than they are today. But recently those funds have backed off, Bannon added, which has depressed pre-seed valuations. For companies that have raised pre-seeds in the last few years, that can make subsequent fundraising more challenging.

“The bar is higher now,” Wu told TechCrunch+. “Investors are starting to ask diligence questions that in 2021 would have been Series A kind of questions.”

Despite investors’ increasing scrutiny at the seed stage, all three investors TechCrunch+ spoke with acknowledge that they don’t expect startups to have it all figured out at this point.

“At the very highest level, you’re investing in people,” Goldberg said. “At the seed stage, even if you have a product and a market, it’s highly likely that you’re going to be changing a bit. The market is going to evolve, and the products are going to evolve. Having a really strong vision and thesis as to where you’re going is critical.”

Bannon concurred, adding that “at seed, most startups don’t have product-market fit, but they should be showing that they have some early signs of it.”

Seed funding milestones and valuations

Beyond expectations about product-market fit, seed round milestones have been ramping up over the years. Recently, the seed round has started to resemble Series A rounds of yesterday.

“I’m advising our portfolio companies that $300,000 to $1 million in ARR is strong to go raise a seed round,” Bannon said. Three years ago, the high end of that was enough to raise a solid Series A.

There are some exceptions, such as deep tech companies or AI startups, she added, where an interesting story or compelling team can raise with less traction in the market. “But I think in general, investors are just looking for more traction than they were a couple of years ago.”

While revenue is not required for deep tech startups, “substantive traction on the business side” is, Wu said. That might include partnerships or collaborations with bigger companies. Investors also want to see a strong proof of concept along with a strong timeline to get to a pilot-scale demonstration. “These are not hard and fast rules, but having clear determination that your thing can work in the real world is where things are at these days,” she said.

Round size at the seed stage is highly variable. In December and January, median deal sizes were $2 million, according to a TechCrunch+ analysis of PitchBook data, with most falling between about $800,000 and $4 million. The median pre-money valuation was $10 million, and the majority spanned $5 million to $20 million.

Those valuations might be a holdover from previous years, Bannon said. She’s been seeing a lot of bridge rounds where the valuations aren’t updated. “You’re taking these inflated prices from two years ago, doing a bridge, and that’s kind of bolstering the market,” she said.

Seed funding market conditions

Though valuations might be inflated, investors remain generally optimistic about this year’s outlook. AI is a driving force. “It’s no longer just buzz. There’s a lot of substance,” Goldberg said. “If you look back, some of the best times to be investing are at the beginning of these major platform shifts and new market areas.”

But that doesn’t mean a rewind to 2021, when deals were closing quickly.

“Everything’s still taking a long time,” Wu said. “Everybody’s bar is higher, and there are also a lot of early stage companies that are raising. That means that seed stage funds can be quite selective and are not necessarily as prone to FOMO as they had been before.”

No matter what the market looks like, though, investors are always looking for new portfolio companies. “We have to make investments both in good markets and in bad markets,” Goldberg said.

The trajectory of today’s seed rounds are somewhat correlated to what happens to startups in the later stages of fundraising. “If the later stage recovers, then the seed stage will just keep going up,” Bannon said.

Advice investors are giving to founders

In previous years, a strong team could show up with a good pitch and walk out with a solid seed round.

“In 2021, you could raise off of one part of your story being really incredible, like an amazing founding team,” Bannon said. “Now it feels like you need an amazing founding team and a compelling business idea and a good market and a little bit of customer traction. I think investors have gotten burned by some of the buzzy teams that have come out and haven’t had something meaningful built.”

Once a startup has those pieces in place, there are other things to consider when planning a seed round. “Time it around an inflection point,” Bannon said. Founders should start laying the groundwork well in advance, she added, so that when revenue or customer growth starts taking hold, they’re ready to pitch.

When founders do walk into a pitch meeting, Goldberg urges them to be prepared not just for the pitch, but also for who is listening. It’s important that founders know the differences between angels, seed funds, and larger venture firms, she said. Also get to know what animates specific investors. The best pitches, she added, are those where the founders clearly show why they’re the best people to bring a product to life.

Goldberg advises founders to manage the fundraising process carefully, to think about who to talk to and when so that when a deal does come together, one firm takes the lead to anchor the round. Without a clear lead, the process can be like “herding cats,” she said, and occupy a lot of a founder’s time. “Make sure that you’re not being overly aggressive, but also aren’t stuck in no-man’s-land where you have a lot of people that aren’t saying no, but not really saying yes.”

In the end, the best way to successfully raise a seed round is to build a business that people want to invest in, Wu said. “Worry less about story, and worry a lot more about making your company a real company. That’s the number one thing.”

What is Bluesky? Everything to know about the app trying to replace Twitter

by with No Comment Tech

Blue sky with clouds illustration, representing Bluesky social

Image Credits: Bryce Durbin / TechCrunch

Is the grass greener on the other side? We’re not sure, but the sky is most certainly bluer. It’s been more than a year since Elon Musk purchased Twitter, now X, leading people to set up shop on alternative platforms. Mastodon, Post, Pebble (which has already shuttered operations) and Spill have been presented as potential replacements, but few aside from Meta’s Threads have achieved the speed of growth Bluesky has reached.

After being invite-only for almost a year, Bluesky is now open to anyone who wants to join. Within a day, Bluesky gained almost 800,000 new users and is slated to break 4 million total signups. Though that number is promising, the network has a lot of catching up to do to compete with Threads’ 130 million monthly active users or even Mastodon’s 1.8 million. And while X users flock to Bluesky anytime Musk makes less than favorable changes, the platform is still able to hit the top of the U.S. App Store.

Here we answer the most common questions about Bluesky social:

What is Bluesky?

Bluesky is a decentralized social app conceptualized by former Twitter CEO Jack Dorsey and developed in parallel with Twitter. The social network has a Twitter-like user interface with algorithmic choice, a federated design and community-specific moderation.

Bluesky is using an open source framework built in-house, the AT Protocol, meaning people outside of the company have transparency into how it is built and what is being developed.

Dorsey introduced the Bluesky project back in 2019 while he was still Twitter CEO. At the time, he said Twitter would be funding a “small independent team of up to five open source architects, engineers, and designers,” charged with building a decentralized standard for social media, with the original goal that Twitter would adopt this standard itself. But that was before Elon Musk bought the platform, so as of late 2022, Bluesky is completely divorced from Twitter. Dorsey has even used Bluesky to express his dismay with Musk’s leadership.

How do you use Bluesky?

Upon signing up, users can create a handle which is then represented as @username.bsky.social as well as a display name that appears more prominent in bold text. If you’re so inclined, you can turn a domain name that you own into your username — so, for example, I’m known on Bluesky as @amanda.omg.lol.

The app itself functions much like a bare-bones Twitter, where you can click a plus button to create a post of 256 characters, which can also include photos. Posts themselves can be replied to, retweeted, liked and, from a three-dot menu, reported, shared via the iOS Share Sheet to other apps or copied as text.

You can search for and follow other individuals, then view their updates in your “Home” timeline. Previously, the Bluesky app would feature popular posts in a “What’s Hot” feed. That feed has since been replaced with an algorithmic and personalized “Discover” feed featuring more than just trending content. 

There is also a “Discover” tab in the bottom center of the app’s navigation, which offers more “who to follow” suggestions and a running feed of recently posted Bluesky updates.

Screenshot of Bluesky menu tab
Image Credits: Natalie Christman

Who’s on Bluesky?

By the beginning of July 2023, when Instagram’s Threads launched, Bluesky topped a million downloads across iOS and Android. Notable figures like Neil Gaiman, Dril and Chelsea Manning have migrated to Bluesky. It’s also home to news organizations and journalists like Bloomberg, The Washington Post and Engadget. With Bluesky now open to the public, it’s possible that other figures could soon make the move over.

Does Bluesky work just like Twitter?

In many ways, yes. However, Bluesky does not yet have DMs or some advanced tools like adding accounts to lists. Additionally, Twitter does not use a decentralized protocol like ActivityPub or AT.

Bluesky was initially kicked off as a project convened by Jack Dorsey in 2019 when he was CEO of Twitter. But the social app has been an independent company since its inception in 2021.

Is Bluesky free?

Yes, and it is now open to the public.

How does Bluesky make money?

Bluesky’s goal is to find another means to sustain its network outside of advertising with paid services, so it can remain free to end users. On July 5, Bluesky announced additional seed round funding and a paid service that provides custom domains for end users who want to have a unique domain as their handle on the service.

Is Bluesky decentralized?

Yes. Bluesky’s team is developing the decentralized AT Protocol, which Bluesky was built atop. In its beta phase, users can only join the bsky.social network, but Bluesky plans to be federated, meaning that endless individually operated communities can exist within the open source network. So, if a developer outside of Bluesky built their own new social app using the AT Protocol, Bluesky users could jump over to the new app and port over their existing followers, handle and data.

“You’ll always have the freedom to choose (and to exit) instead of being held to the whims of private companies or black box algorithms. And wherever you go, your friends and relationships will be there too,” a Bluesky blog post explained.

Is Bluesky secure?

In October 2023, Bluesky added email verification as part of a larger effort to improve account security and authentication on the network. The addition is an important step forward in terms of making Bluesky more competitive with larger networks like X, which have more robust security controls. In December 2023, Bluesky allowed users to opt out of a change that would expose their posts to the public web following backlash from users. 

Is Bluesky customizable?

Yes. In May 2023, Bluesky released custom algorithms, which it calls “custom feeds.” Custom feeds allow users to subscribe to multiple different algorithms that showcase different kinds of posts a user may want to see. You can pin custom feeds that will show up at the top of your timeline as different tabs to pick from. The feeds you pin, or save, are located under the “My Feeds” menu in the app’s sidebar.

In March 2024,​​ the company announced “AT Protocol Grants,” a new program that will dole out small grants to developers in order to foster growth and customization. One of the recipients, SkyFeed, is a custom tool that lets anyone build their own feeds using a graphical user interface. 

Is Bluesky on iOS and Android?

Yes. Bluesky rolled out to Android users on April 20 and was initially launched to iOS users in late February. Users can access Bluesky on the web here.

Are Bluesky posts really called ‘skeets?’

There is technically no name for posts, but internet users have adopted the name “skeets,” a portmanteau of “tweet” and “sky.” Users still widely refer to posts as “skeets,” despite protests from Bluesky CEO Jay Graber and others who don’t find the slang for semen amusing.

How does Bluesky tackle misinformation?

After an October 2023 update, the app will now warn users of misleading links by flagging them. If links shared in users’ posts don’t match their text, the app will offer a “possibly misleading” warning to the user to alert them that the link may be directing them somewhere they don’t want to go.

Image Credits: Bluesky on GitHub

Has Bluesky had any controversies?

Bluesky has been embattled with moderation issues since its first launch. The app has been accused of failing to protect its marginalized users and failing to moderate racist content. Following a controversy about the app allowing racial slurs in account handles, frustrated users initiated a “posting strike,” where they refused to engage with the platform until it established guardrails to flag slurs and other offensive terms in usernames.

What moderation features does Bluesky have?

A December 2023 post from the Bluesky Safety account announced a large batch of moderation updates. 

Bluesky is rolling out “more advanced automated tooling” designed to flag content that violates its Community Guidelines that will then be reviewed by the app’s moderation team. 

Bluesky launched moderation features similar to ones on X, including user lists and moderation lists, the latter of which can be used to mute or block many users at once. The app is also developing a feature that lets users limit who can reply to posts.

Some Bluesky users are still advocating for the ability to set their accounts to private — a feature they have an increased need for after Bluesky announced it would launch a public web interface. 

In March 2024, the company launched Ozone, a tool that lets users create and run their own independent moderation services that will give users “unprecedented control” over their social media experience.

What’s the difference between Bluesky and Mastodon?

Though Bluesky’s architecture is similar to Mastodon’s, many users have found Bluesky to be more intuitive, while Mastodon can come off as inaccessible: Choosing which instance to join feels like an impossible task on Mastodon, and longtime users are very defensive about their established posting norms, which can make joining the conversation intimidating. To remain competitive, Mastodon recently simplified its sign-up flow, making mastodon.social the default server for new users.

However, the launch of federation will make it work more similarly to Mastodon in that users can pick and choose which servers to join and move their accounts around at will. 

Who owns Bluesky?

Though Jack Dorsey funded Bluesky and sits on the company’s board, he is not involved in day-to-day development. The CEO of Bluesky is Jay Graber, who previously worked as a software engineer for the cryptocurrency Zcash, then founded an event-planning site called Happening.

If you have more FAQs about Bluesky not covered here, leave us a comment below. 

Apple vs US antitrust lawsuit: Everything we know so far on the DOJ's iPhone case

by with No Comment Tech

Image Credits: Bryce Durbin/TechCrunch

Apple’s antitrust scrutiny has reached a fever pitch. The U.S. Department of Justice announced Thursday that it filed a lawsuit accusing the company of behaving like a monopoly in locking in iPhone customers and limiting competitors building hardware and software. The lawsuit, which comes on the heels of significant antitrust cases against Apple outside the U.S., is a wide-ranging and complicated affair, but we’re covering the ins and outs of the DOJ’s case, the industry’s response and all the ongoing implications for companies and customers.

We’ll be updating this page as the Apple antitrust case evolves, but keep in mind that there will be little settled in the short term. Experts estimate a three-to-five-year timeline for any resolution for the case.

The DOJ’s claims against Apple

If you want to dive into legal docs immediately, you can read the DOJ’s lawsuit right here. But for the rest of us, there are five categories that the complaint identifies as areas in which Apple actively suppressed competition.

“Super” apps: These are applications that contain numerous functions within a single app. This should ring a bell for anyone following Elon Musk’s “everything app” aspirations for X, and the DOJ claims Apple is inhibiting their success to increase dependence on the iPhone.

Messaging apps: The blue bubble, green bubble effect is specifically cited by the DOJ as a factor discouraging iPhone users from adopting a competitor device. “This effect is particularly powerful for certain demographics, like teenagers — where the iPhone’s share is 85 percent, according to one survey,” the DOJ said in the lawsuit.

Cloud streaming gaming apps: The DOJ lawsuit highlights Apple’s alleged opposition to cloud-based gaming, claiming its actions are to prevent consumers from playing games “without the need for users to purchase powerful, expensive hardware.”

Digital wallets: Though the 0.15% fee Apple takes for all transactions made through Apple Pay is a fraction of the company’s total revenue, the DOJ alleges that the ubiquity of Apple Pay within its mobile ecosystem means it has “complete control” over users’ NFC payments and that it hinders competitors.

Smartwatch cross-platform compatibility: This DOJ claim is straightforward. By limiting the functionality of Apple Watches with non-iPhone devices, the lawsuit claims “it becomes more costly for that user to purchase a different kind of smartphone.”

Apple, and the wider industry’s, response

Apple issued an extensive series of rebuttals to the DOJ’s claims Thursday, which you can check out in full detail right here. The core of Apple’s argument is that regulators are selectively picking metrics that make Apple’s strength in the smartphone market seem more dominant than it actually is, in their view. And in regulating the behaviors that the DOJ claims are monopolistic, Apple’s competitive advantage in the market would be diminished and iPhone customers negatively impacted in the process.

“This lawsuit threatens who we are and the principles that set Apple products apart in fiercely competitive markets. If successful, it would hinder our ability to create the kind of technology people expect from Apple — where hardware, software, and services intersect,” Apple said in a statement provided to TechCrunch

App makers are less critical of the DOJ’s case, with the Coalition for App Fairness (CAF) voicing strong support for the DOJ’s regulatory action, which comes as no surprise given several of its members, like Epic Games and Spotify, have already had public disputes with Apple on its App Store practices.

“The DOJ complaint details Apple’s long history of illegal conduct — abusing their App Store guidelines and developer agreements to increase prices, extract exorbitant fees, degrade user experiences, and choke off competition,” CAF Executive Director Rick VanMeter said in a statement Thursday. “The DOJ joins regulators around the world, who have recognized the many harms of Apple’s abusive behavior and are working to address it.”

What could this mean for iPhone users, and what’s next?

In the immediate term, not much. The release of the lawsuit, and the ensuing back-and-forth between Apple and the DOJ, was a flurry of activity that will take years to settle. The DOJ’s antitrust case against Google, which was filed back in 2020, went to trial last year and could still take a couple more years to reach a conclusion.

What you shouldn’t expect is for the present to play out like the past. Though the DOJ cites the successful antitrust prosecution against Microsoft in the 1990s, there are many distinctions between the two cases, mostly notably a gap between how easily defined Microsoft’s market dominance was compared to Apple’s current status quo.

For more on Apple’s antitrust lawsuit, check here:

Apple sued by DOJ over iPhone monopoly claimsDOJ claims green bubbles are an issue in Apple iPhone antitrust lawsuitWhy Apple’s antitrust lawsuit could be a silver lining for Epic GamesApple’s iPhone is not a monopoly like Windows was a monopoly Epic, Spotify, Deezer, Match Group and others applaud DOJ’s Apple lawsuitDOJ calls out Apple for breaking iMessage-on-Android solution, BeeperHere’s what the DOJ suit could mean for Apple WatchDOJ says Apple’s ‘complete control’ over tap-to-pay transactions stops innovationApple slams DOJ case as misguided attempt to turn iPhone into AndroidDOJ’s Apple antitrust case neatly aligns with EU on one key point: NFC and mobile paymentsThe DOJ’s case against Apple adds to a growing pile of antitrust problems for Cupertino

illustration featuring Google's Bard logo

Google Gemini: Everything you need to know about the new generative AI platform

by with No Comment Tech

illustration featuring Google's Bard logo

Image Credits: TechCrunch

Google’s trying to make waves with Gemini, its flagship suite of generative AI models, apps and services.

So what is Gemini? How can you use it? And how does it stack up to the competition?

To make it easier to keep up with the latest Gemini developments, we’ve put together this handy guide, which we’ll keep updated as new Gemini models, features and news about Google’s plans for Gemini are released.

What is Gemini?

Gemini is Google’s long-promised, next-gen GenAI model family, developed by Google’s AI research labs DeepMind and Google Research. It comes in three flavors:

Gemini Ultra, the most performant Gemini model.Gemini Pro, a “lite” Gemini model.Gemini Nano, a smaller “distilled” model that runs on mobile devices like the Pixel 8 Pro.

All Gemini models were trained to be “natively multimodal” — in other words, able to work with and use more than just words. They were pretrained and fine-tuned on a variety of audio, images and videos, a large set of codebases and text in different languages.

This sets Gemini apart from models such as Google’s own LaMDA, which was trained exclusively on text data. LaMDA can’t understand or generate anything other than text (e.g., essays, email drafts), but that isn’t the case with Gemini models.

What’s the difference between the Gemini apps and Gemini models?

Google's Bard
Image Credits: Google

Google, proving once again that it lacks a knack for branding, didn’t make it clear from the outset that Gemini is separate and distinct from the Gemini apps on the web and mobile (formerly Bard). The Gemini apps are simply an interface through which certain Gemini models can be accessed — think of it as a client for Google’s GenAI.

Incidentally, the Gemini apps and models are also totally independent from Imagen 2, Google’s text-to-image model that’s available in some of the company’s dev tools and environments.

What can Gemini do?

Because the Gemini models are multimodal, they can in theory perform a range of multimodal tasks, from transcribing speech to captioning images and videos to generating artwork. Some of these capabilities have reached the product stage yet (more on that later), and Google’s promising all of them — and more — at some point in the not-too-distant future.

Of course, it’s a bit hard to take the company at its word.

Google seriously underdelivered with the original Bard launch. And more recently it ruffled feathers with a video purporting to show Gemini’s capabilities that turned out to have been heavily doctored and was more or less aspirational.

Google’s best Gemini demo was faked

Still, assuming Google is being more or less truthful with its claims, here’s what the different tiers of Gemini will be able to do once they reach their full potential:

Gemini Ultra

Google says that Gemini Ultra — thanks to its multimodality — can be used to help with things like physics homework, solving problems step-by-step on a worksheet and pointing out possible mistakes in already filled-in answers.

Gemini Ultra can also be applied to tasks such as identifying scientific papers relevant to a particular problem, Google says — extracting information from those papers and “updating” a chart from one by generating the formulas necessary to re-create the chart with more recent data.

Gemini Ultra technically supports image generation, as alluded to earlier. But that capability hasn’t made its way into the productized version of the model yet — perhaps because the mechanism is more complex than how apps such as ChatGPT generate images. Rather than feed prompts to an image generator (like DALL-E 3, in ChatGPT’s case), Gemini outputs images “natively,” without an intermediary step.

Gemini Ultra is available as an API through Vertex AI, Google’s fully managed AI developer platform, and AI Studio, Google’s web-based tool for app and platform developers. It also powers the Gemini apps — but not for free. Access to Gemini Ultra through what Google calls Gemini Advanced requires subscribing to the Google One AI Premium Plan, priced at $20 per month.

The AI Premium Plan also connects Gemini to your wider Google Workspace account — think emails in Gmail, documents in Docs, presentations in Sheets and Google Meet recordings. That’s useful for, say, summarizing emails or having Gemini capture notes during a video call.

Gemini Pro

Google says that Gemini Pro is an improvement over LaMDA in its reasoning, planning and understanding capabilities.

An independent study by Carnegie Mellon and BerriAI researchers found that the initial version of Gemini Pro was indeed better than OpenAI’s GPT-3.5 at handling longer and more complex reasoning chains. But the study also found that, like all large language models, this version of Gemini Pro particularly struggled with mathematics problems involving several digits, and users found examples of bad reasoning and obvious mistakes.

Early impressions of Google’s Gemini aren’t great

Google promised remedies, though — and the first arrived in the form of Gemini 1.5 Pro.

Designed to be a drop-in replacement, Gemini 1.5 Pro is improved in a number of areas compared with its predecessor, perhaps most significantly in the amount of data that it can process. Gemini 1.5 Pro can take in ~700,000 words, or ~30,000 lines of code — 35x the amount Gemini 1.0 Pro can handle. And — the model being multimodal — it’s not limited to text. Gemini 1.5 Pro can analyze up to 11 hours of audio or an hour of video in a variety of different languages, albeit slowly (e.g., searching for a scene in a one-hour video takes 30 seconds to a minute of processing).

Gemini 1.5 Pro entered public preview on Vertex AI in April.

An additional endpoint, Gemini Pro Vision, can process text and imagery — including photos and video — and output text along the lines of OpenAI’s GPT-4 with Vision model.

Gemini
Using Gemini Pro in Vertex AI. Image Credits: Gemini

Within Vertex AI, developers can customize Gemini Pro to specific contexts and use cases using a fine-tuning or “grounding” process. Gemini Pro can also be connected to external, third-party APIs to perform particular actions.

Google brings Gemini Pro to Vertex AI

In AI Studio, there’s workflows for creating structured chat prompts using Gemini Pro. Developers have access to both Gemini Pro and the Gemini Pro Vision endpoints, and they can adjust the model temperature to control the output’s creative range and provide examples to give tone and style instructions — and also tune the safety settings.

Gemini Nano

Gemini Nano is a much smaller version of the Gemini Pro and Ultra models, and it’s efficient enough to run directly on (some) phones instead of sending the task to a server somewhere. So far, it powers a couple of features on the Pixel 8 Pro, Pixel 8 and Samsung Galaxy S24, including Summarize in Recorder and Smart Reply in Gboard.

The Recorder app, which lets users push a button to record and transcribe audio, includes a Gemini-powered summary of your recorded conversations, interviews, presentations and other snippets. Users get these summaries even if they don’t have a signal or Wi-Fi connection available — and in a nod to privacy, no data leaves their phone in the process.

Gemini Nano is also in Gboard, Google’s keyboard app. There, it powers a feature called Smart Reply, which helps to suggest the next thing you’ll want to say when having a conversation in a messaging app. The feature initially only works with WhatsApp but will come to more apps over time, Google says.

And in the Google Messages app on supported devices, Nano enables Magic Compose, which can craft messages in styles like “excited,” “formal” and “lyrical.”

Is Gemini better than OpenAI’s GPT-4?

Google has several times touted Gemini’s superiority on benchmarks, claiming that Gemini Ultra exceeds current state-of-the-art results on “30 of the 32 widely used academic benchmarks used in large language model research and development.” The company says that Gemini 1.5 Pro, meanwhile, is more capable at tasks like summarizing content, brainstorming and writing than Gemini Ultra in some scenarios; presumably this will change with the release of the next Ultra model.

But leaving aside the question of whether benchmarks really indicate a better model, the scores Google points to appear to be only marginally better than OpenAI’s corresponding models. And — as mentioned earlier — some early impressions haven’t been great, with users and academics pointing out that the older version of Gemini Pro tends to get basic facts wrong, struggles with translations and gives poor coding suggestions.

How much does Gemini cost?

Gemini 1.5 Pro is free to use in the Gemini apps and, for now, AI Studio and Vertex AI.

Once Gemini 1.5 Pro exits preview in Vertex, however, the model will cost $0.0025 per character while output will cost $0.00005 per character. Vertex customers pay per 1,000 characters (about 140 to 250 words) and, in the case of models like Gemini Pro Vision, per image ($0.0025).

Let’s assume a 500-word article contains 2,000 characters. Summarizing that article with Gemini 1.5 Pro would cost $5. Meanwhile, generating an article of a similar length would cost $0.1.

Ultra pricing has yet to be announced.

Where can you try Gemini?

Gemini Pro

The easiest place to experience Gemini Pro is in the Gemini apps. Pro and Ultra are answering queries in a range of languages.

Gemini Pro and Ultra are also accessible in preview in Vertex AI via an API. The API is free to use “within limits” for the time being and supports certain regions, including Europe, as well as features like chat functionality and filtering.

Elsewhere, Gemini Pro and Ultra can be found in AI Studio. Using the service, developers can iterate prompts and Gemini-based chatbots and then get API keys to use them in their apps — or export the code to a more fully featured IDE.

Code Assist (formerly Duet AI for Developers), Google’s suite of AI-powered assistance tools for code completion and generation, is using Gemini models. Developers can perform “large-scale” changes across codebases, for example updating cross-file dependencies and reviewing large chunks of code.

Google’s brought Gemini models to its dev tools for Chrome and Firebase mobile dev platform, and its database creation and management tools. And it’s launched new security products underpinned by Gemini, like Gemini in Threat Intelligence, a component of Google’s Mandiant cybersecurity platform that can analyze large portions of potentially malicious code and let users perform natural language searches for ongoing threats or indicators of compromise.

Gemini Nano

Gemini Nano is on the Pixel 8 Pro, Pixel 8 and Samsung Galaxy S24 — and will come to other devices in the future. Developers interested in incorporating the model into their Android apps can sign up for a sneak peek.

Is Gemini coming to the iPhone?

It might! Apple and Google are reportedly in talks to put Gemini to use for a number of features to be included in an upcoming iOS update later this year. Nothing’s definitive, as Apple is also reportedly in talks with OpenAI and has been working on developing its own GenAI capabilities.

This post was originally published Feb. 16, 2024 and has since been updated to include new information about Gemini and Google’s plans for it.