OpenAI has closed a yawning gap in its ChatGPT subscription pricing with a new $100 per month Pro plan that slots between the $20 per month Plus plan and $200 per month Pro plan. Offering five times more Codex than the $20 option, it appears designed to challenge Anthropic's $100 per month Claude option. "Compared with Claude Code, Codex delivers more coding capacity per dollar across paid tiers," an OpenAI spokesperson told TechCrunch.
So what's the difference between OpenAI's two Pro plans? The $200 version does offer four times the Codex. However, you get the same advanced tools and models with $100 plan, according to OpenAI's product page. To encourage users to jump in, it will offer double the Codex for a limited time, or 10 times what you get with the Plus plan.
Users have been screaming for such a plan for a while now, according to posts on OpenAI's developer community forums. "The Plus plan will continue to be the best offer at $20 for steady, day-to-day usage of Codex, and the new $100 Pro tier offers a more accessible upgrade path for heavier daily use," OpenAI said in a post on X.
With the launch of GPT 5.2 late last year and GPT-5.3-Codex in February, OpenAI significantly boosted the speed and reasoning capabilities of Codex, giving developers a tough choice between ChatGPT and Claude Opus. However, the sticking point for many power users was ChatGPT's $200 per month price — so OpenAI no doubt hopes the new plan will convince those on the fence to switch.
This article originally appeared on Engadget at https://www.engadget.com/ai/openai-has-a-new-100-chatgpt-pro-plan-to-better-match-up-with-claude-062705626.html?src=rss
This year at CES 2026 everybody was pretty confused about the new "Micro RGB" and "RGB Mini LED" TVs that use similar technology but carry different names. Now, Sony has come up with another label for its own Mini LED TVs with RGB backlighting: True RGB. The idea is to emphasize that the individual red, green and blue LED backlights allow for "purer color, greater brightness, and the largest color volume ever achieved in Sony’s home TV history," the company said.
To be clear, this is not some new technology that Sony just came up with — it's the same Micro RGB tech we saw earlier this year from Samsung, LG, HiSense and others. These TVs use pure red, green and blue LED backlights along with an LCD layer (rather than solid blue LEDs and quantum dots like Mini LED TVs) to produce the final picture. This display tech is supposed to deliver better color accuracy and more brightness than regular Mini LED TVs. (It's not the same as OLED tech, in which each pixel acts as a light source.)
Sony's True RGB backlight tech (right) compared to current Mini LED TVs
Sony
However, Sony says that the way it processes the image makes its True RGB TVs stand out from rivals. To control the LEDs more precisely, it borrowed algorithms from its wildly expensive professional reference monitors. That supposedly allows for more precise color control and higher brightness that allow movies and series to look more like the creators intended. It also reduces the "blooming" that occurs when light leaks into neighboring pixels, while improving color accuracy when viewing the TVs from an angle.
Every TV maker claims to have the best technology, but Sony has a lot of credibility due to its history with cinema cameras, Hollywood productions and reference monitors. We'll have to wait until spring this year to see the new Bravia True RGB TVs for ourselves, but prior to that, the company has promised to release "additional details" about them in the near future.
This article originally appeared on Engadget at https://www.engadget.com/home/home-theater/sony-teases-its-next-gen-true-rgb-mini-led-tv-technology-064732375.html?src=rss
Apple has run into "more issues than expected" with its foldable iPhone that may set back its release, according to Nikkei. The engineering problems reportedly cropped up during the device's early test production phase and may delay first shipments by months, according to multiple sources briefed on the matter. However, a separate report in Bloomberg refutes the gist of Nikkei's claims.
"The current situation could put the mass production timeline at risk," one of Nikkei's sources said. "April will mark a crucial stage of the engineering verification test, and this month till early may is extremely critical." Component suppliers have supposedly been notified that the foldable iPhone's production schedule will be delayed, and Apple is working to address the problems.
In a separate article inBloomberg, however, the usually reliable Apple reporter Mark Gurman wrote that the device is still on track for a September 2026 release. "The company is scheduled to introduce the foldable model in September alongside the iPhone 18 Pro and Pro Max, said the people, who asked not to be identified because the plans haven’t been announced," according to the report.
The complexity of the foldable iPhone's display and other components may "limit initial supply for several weeks," but Apple plans to put it on sale at the same time or shortly after the other non-foldable devices, Gurman wrote. That said, the timing isn't final and issues could still crop up ahead of production. Engadget has reached out to Apple, but so far the company has declined to comment on either the Bloomberg or Nikkei reports.
A foldable iPhone has been rumored since 2017, and Apple's biggest rival, Samsung, released its first one back in 2019. According to multiple sources, Apple was aiming to launch its debut foldable iPhone in fall 2026 alongside the iPhone 18. However, as we detailed in an explainer last month, "the project could slip into 2027 if Apple runs into manufacturing or durability issues, particularly around the hinge or display."
Apple was reportedly prioritizing the foldable iPhone and other premium models for its September event this year due to constrained supplies of components like memory chips. "Apple and the supply chain are working under a pressured timeline and the current solutions are not enough to completely solve the engineering challenge... more time is needed," Nikkei's source stated.
The problems reportedly arose during Apple's production verification tests. That's the fourth of six steps the company's new products must go through before shipping, prior to the key pilot production and mass production phases. Since the foldable would be an all-new design, it would likely need to pass each stage with flying colors before proceeding to the next.
Though likely to account for less than 10 percent of iPhone production, the foldable will be a key product for Apple designed to boost interest in iPhones across its range. Apple reportedly plans to produce seven to eight million of the devices initially, Nikkei reported. Apple has yet to announce the device.
Update April 7, 2026 at 3:02 PM ET: The article has been updated with information from Bloomberg's Mark Gurman stating that the foldable iPhone is still on schedule for a September 2026 launch.
This article originally appeared on Engadget at https://www.engadget.com/mobile/smartphones/apples-foldable-iphone-may-be-delayed-due-to-engineering-snags-073813890.html?src=rss
Earlier this year, Amazon threatened to cut US Postal Service deliveries by as much as two thirds. Now, the parties have reached tentative a deal that will see USPS deliveries reduced by 20 percent, The Wall Street Journal reported. While not as drastic as first menaced, the reduced volume will still deal a financial blow to the USPS.
"We’re pleased to have reached a new agreement with USPS that furthers our longstanding partnership and will let us continue supporting our customers and communities together," an Amazon spokesperson told the WSJ.
Amazon is the USPS's largest customer, accounting for 15 percent of its volume and $6 billion in revenue. A two-thirds cut would have been a disaster for the USPS, but a 20 percent reduction could result in more than $1 billion in lost revenue nonetheless. Amazon would have needed to scramble as well, as it relies heavily on the post office for rural and last-mile deliveries.
Amazon's contract with the USPS was set to expire in September 2026, and in October Amazon said it wanted to strike a deal by December 2025. However, the USPS abruptly pulled out of negotiations, according to Amazon, and implemented a new bidding process for last-mile deliveries. "Our goal was to increase our volumes with USPS, not reduce them — until USPS abruptly walked away at the eleventh hour in December," Amazon said at the time.
Amazon was reportedly considering expanding its own delivery network if the USPS deal fell through, though the company may have started those rumors itself to prod negotiations. The Postal Service decided to re-engage with Amazon after bids from several Amazon rivals fell short of its volume and revenue expectations, according to the WSJ's sources. The new agreement is still subject to approval by the federal Postal Regulatory Commission.
This article originally appeared on Engadget at https://www.engadget.com/general/amazons-new-usps-deal-will-see-postal-deliveries-cut-by-20-percent-054608944.html?src=rss
Ahead of its premiere, Dave Filoni has revealed that the Star Wars animated series Maul: Shadow Lord will return for a second season. The Lucasfilm co-president revealed that season 2 is already in the works, telling Esquire that "at the end of the day, people like that character."
Filoni didn't reveal any other details about the plot or release date for season 2. However, the news isn't a great surprise given Lucasfilm's past history with its animated series — The Clone Wars ran seven seasons, Star Wars Rebels four seasons, Star Wars Resistance two seasons and Star Wars: The Bad Batch three seasons.
Maul: Shadow Lord explores the Zebrak Sith Lord's story about a year after the time of the Clone Wars. Season 1's 10 seasons will stream twice a week on Disney+ starting on April 6 and run through May 6. It covers Maul's plot to rebuild his criminal syndicate "on a planet untouched by the Empire," according to Lucasfilm. "There, he crosses paths with a disillusioned young Jedi Padawan who may just be the apprentice he is seeking to aid him in his relentless pursuit for revenge."
This article originally appeared on Engadget at https://www.engadget.com/entertainment/streaming/star-wars-animated-series-maul-shadow-lord-season-2-confirmed-054036065.html?src=rss
Every now and then, I test a gadget so wild that I can’t believe a company actually made it. Soundcore’s $5,000 Nebula X1 Pro projector is the embodiment of that: an ultra bright projector and a 400-watt Dolby Atmos 7.1 speaker system combined in a massive enclosure. With a fast and flexible setup, it lets you screen movies or watch sports nearly anywhere.
It’s not just a projector crammed into a big speaker system, though. Everything is elegantly integrated and setup is nearly automatic, thanks to the clever design and motorization. The weight and price are the biggest strikes against it, but if you can afford it, and love hosting movie nights, the Nebula X1 Pro is one of the coolest devices you can buy.
Design
Made by Anker sub-brand Soundcore, the Nebula X1 Pro has a professional-looking enclosure housing a Nebula X1 laser projector and five speakers — a subwoofer, two front satellites and two rear satellites. With all that crammed in, the projector is big and heavy at 30 inches high and 72 pounds. Fortunately, it has a pair of wheels on the back and a telescoping handle so it’s easy to roll from room to room or dolly outside. Good luck carrying it up a set of stairs or unloading it from a vehicle by yourself, though.
Steve Dent for Engadget
Soundcore made the Nebula X1 Pro as outdoor-friendly as possible, with IP43 and IP54 ratings on the body and speakers, respectively, to withstand short periods of rain. If you want to use it away from home, the company sells optional kits with a 200-inch inflatable screen or an Anker Solix C1000 battery that can power it for several hours.
The four wireless satellite speakers have seven horizontal and four overhead channels, and the two-speaker subwoofer is inside the main enclosure. The rear speakers pop out of a spring-loaded storage dock with a light press, same for the telescoping speaker legs.
The front satellite speaker docking system is even more slick. To release them, you press a button on top and they fold out of the side via a motorized system. You can either leave them there or detach them at the touch of a button. All speakers can be charged externally over USB-C or inside their docks. They have eight hours of battery life, though I found the bigger front ones held a charge for slightly longer than that.
The X1 Pro also includes a pair of high-quality Soundcore wireless microphones for DJ-ing or karaoke, tucked under the top panel. Those feature AI vocal removal from songs, one-touch reverb and 40 hours of battery life — everything you need for a karaoke party.
There’s only a single HDMI 2.1 port at the back (which is odd considering that the Nebula X1 has two) with eARC support for Dolby Atmos sound. It also comes with two USB-C ports for external file playback, charging and a PC connection. Lastly, the power cable is retractable, which is another nice design touch.
Features
Steve Dent for Engadget
I also reviewed the Nebula X1 projector that's inside the X1 Pro, but here’s a summary in case you missed it. The projector uses Soundcore’s proprietary “LaserForge 2.0” liquid-cooled, triple-laser engine that beams a bright, color-accurate image with very little fan noise (26 db). It promises high native contrast thanks to the 6-blade dynamic iris and NebulaMaster 2.0 image engine. The 0.9:1 to 1.5:1 optical zoom lens allows for flexible installation and employs 14 high-quality, long-lasting glass elements.
The X1 Pro uses the same 0.47-inch DLP chip found in many other projectors (and not the bigger, better 0.67-inch chip coming soon in XGIMI’s Titan Noir). The lasers are beamed through a color phosphor wheel twice to achieve excellent 90 percent color and brightness uniformity across the screen.
The projector’s motorized gimbal tilts 25 degrees upward so you can position it well below the screen. The “spatial adaptation” feature scans the projection area then beams the final image to precisely fit the screen or wall. It worked nicely for me, though overhead lights or other obstacles can throw it off. The projector can adapt to ambient light and the wall color, and another function called Spatial Recall lets you save all your settings for later.
Once I detached them, the speakers paired automatically to the X1 Pro over 5.8Ghz Wi-Fi with no difficulty. I placed them around the room to maximize soundstage, then the Nebula X1 Pro’s “Flexwave” tech used a built-in four-mic array to detect their positions and calibrate the audio. I was seated off to the side, so I used the “smart sweet spot” feature to drag the center point toward my position for optimal sound balance.
Google TV is included, offering a large library of streaming apps and an easy-to-use projector control interface via the included remote (tucked into the top so you hopefully won’t lose it). You get Netflix’s official app with support for 4K Dolby Vision and Dolby Atmos, plus the X1 Pro has Chromecast support and Google Assistant for voice control. The interface can occasionally be sluggish, though Soundcore has improved its latency since I tested the Nebula X1.
Image quality
Steve Dent for Engadget
Even after testing other high-end projectors including Valerion’s VisionMaster Max, the Soundcore Nebula X1 Pro is still the brightest and sharpest I’ve seen. The company’s luminosity claim is accurate; I measured 3,514 ANSI lumens in “Standard” mode from the center of the screen and 3,310 in the cinematic “NebulaMaster” mode. It can output a whopping 4,175 lumens in Conference mode, albeit with a heavy blue color cast.
That brightness allowed me to comfortably watch content on a sunny day with the shades up. The X1 Pro also offers high dynamic contrast up to 56,000:1, aided by the automatic iris and NebulaMaster image engine, which also keeps the image from washing out in daytime conditions.
When used in more ideal dark conditions, the image was bright, sharp and incredibly color accurate. Soundcore claims 110 percent coverage of the challenging BT.2020 HDR color space (with a Delta E less than 0.8), putting the X1 Pro in elite company with a few select models from Samsung, Hisense and a few others. I measured around 94 percent BT.2020 coverage in ISF mode, which falls short of the company’s claim but is still impressive.
The high color accuracy meant that the TV series and movies I watched like Iron Man 2, Dune 2, Andor and F1 looked beautifully cinematic. If the colors aren’t quite to your liking, you can make fine adjustments manually. Like other 4K projectors with a 0.47-inch DLP chip, the X1 has a slight amount of light spill around the edge of the screen, but it’s only noticeable when the projected image is particularly dark.
With HDMI 2.1 the Nebula X1 Pro supports 4K 120 fps sources, but can only display 4K at 60 fps. Because of that, and the relatively high input lag, it’s not ideal for gaming.
Audio
Steve Dent for Engadget
The 400-watt audio setup is what elevates the Nebula X1 Pro above its rivals. Thanks to their Wi-Fi connectivity, the satellites have a latency of just 25 milliseconds, compared to 150 milliseconds or more for typical Bluetooth speakers. That keeps sound and picture perfectly synced, something that can be a problem with other wireless speaker setups.
The X1’s two internal subwoofers can pump out sound as low as 38Hz at up to 87 decibels. That allowed for the loud and punchy (but not boomy) bass I love for action movies like Spider-Man: No Way Home and Once Upon a Time in Hollywood. At the same time, that bass is clear and subtle for less bombastic films like Eternal Sunshine of the Spotless Mind. To avoid vibrating the projector, the subwoofer is mounted on a suspension system, and I found it didn’t affect the picture even during loud scenes.
With the four wireless speakers spread around a big room, I got an outstanding soundstage with Dolby Atmos-supported content including Star Wars: A New Hope and The White Lotus:Season 3. The speakers delivered crisp and accurate highs, while the dedicated front voice drivers let me hear even soft dialogue, though midrange sound could occasionally be a bit tinny. It faithfully reproduced tricky film soundtracks like Lord of the Rings: The Two Towers and did justice to the industrial metal and symphonic music in The Matrix. Yes, you’d get better sound from a dedicated high-end 7.1 Dolby system, but with far more setup hassle and zero portability.
Wrap-up
Steve Dent for Engadget
Soundcore’s Nebula X1 Pro is a home theater marvel that’s so well-designed almost anyone can set it up. By integrating one of the best triple-laser projectors with a 400-watt Dolby Atmos 7.1.4 surround system, and then putting all of that on wheels, you can enjoy an immersive cinema experience nearly anywhere.
This Nebula X1 Pro has no true rivals, but competitors with similar projectors (but no sound systems) include the Valerion VisionMaster Max and XGIMI Horizon 20 Max, both triple-laser systems with comparable brightness and color accuracy. Once you add an audio surround system, though, you’ll be spending the same amount and won’t get the X1 Pro’s convenience and portability.
The catch, of course, is the $5,000 price. However, if you have the money and want the ultimate home theater experience that’s portable and easy to use, Soundcore’s beastly Nebula X1 Pro is actually a good deal.
This article originally appeared on Engadget at https://www.engadget.com/home/home-theater/soundcore-nebula-x1-pro-review-the-king-of-party-projectors-010018484.html?src=rss
Google's $20 per month AI Pro plan, which includes Gemini, Veo and Nano Banana, got a big storage boost and some other new perks. Users of the plan (also available for $200 per year) will see their cloud space jump from 2TB to 5TB at no extra cost. That extra storage can be used not only for AI but also Gmail, Google Drive and Google Photos backups.
Gemini can now pull context from Gmail and the web for Drive, Docs, Slides and Sheets, provide summaries for your Gmail inbox and proofread emails before you send them. It's also introducing additional agentic help with Chrome auto browse "that handles those tedious, multi-step chores — like planning a trip or filling out forms," Google VP Shimrit Ben-Yair wrote on X.
Finally, Google announced that it's bundling its Home Premium subscription into AI Pro, a perk that usually costs $10 per month by itself. The storage and extra features are now available for new and existing subscribers. You may not see the benefits appear in your plan yet but it's definitely not an April Fool's joke, Ben-Yair assured X commenters.
This article originally appeared on Engadget at https://www.engadget.com/ai/googles-20-per-month-ai-pro-plan-just-got-a-big-storage-boost-044502621.html?src=rss
After releasing DLSS 4.5's Super Resolution feature earlier this year, NVIDIA has released an update with DLSS 4.5 features designed to boost frame rates on RTX 50 series cards. Those include DLSS 4.5 Dynamic Multi Frame Generation and DLSS 4.5 Multi Frame Generation 6X. With those, NVIDIA is promising the "smoothest path-traced gaming yet" to unlock the potential for high-refresh 4K 240Hz OLED gaming displays, or 1080p and 1440p monitors at 360Hz and beyond.
The first feature, DLSS 4.5 Dynamic Multi Frame Generation, is like an "automatic transmission" for your RTX 50 series card, NVIDIA said. Rather than multiplying the frame rate by a fixed amount, the AI-powered feature changes it dynamically to strike a balance between refresh rate, image quality and responsiveness. To optimize computing power, it ensures that a game's frame rate doesn't exceed your monitor's native refresh rate, so you won't play at 240 fps on a 120Hz monitor.
The other key feature, Multi Frame Generation 6X, is designed to deliver even higher levels of performance. Based on NVIDIA's second-gen transformer model, along with frame pacing and image quality improvements, the feature boosts the maximum multiplier to 6X, generating up to five extra frames for every natively rendered frame on GeForce RTX 50 series GPUs. That will boost 4K frame rates up to 35 percent "with minimal impact to responsiveness," NVIDIA wrote.
As with Super Resolution, the native frame rate of a game is not accelerated by these features. Rather, the DLSS 4.5 feature uses AI to create interpolated intermediate frames the machine "thinks" should be there. Normally this works fine, but the scheme can create odd artifacts in certain types of scenes, particularly with fine details like rain, hair and phone wires.
Still, the tech allows for smoother gaming with a minimal increase on your GPU's burden. It's now available for NVIDIA RTX 50 series cards, and as NVIDIA announced earlier, there are a number of games supporting the MFG features. Those include 007 First Light (May 27th), CONTROL Resonant, Directive 8020 (May 12th) and Tides of Annihilation.
This article originally appeared on Engadget at https://www.engadget.com/gaming/pc/nvidias-dlss-45-multi-frame-generation-tech-is-now-available-to-boost-your-hz-130035249.html?src=rss
After achieving action camera success, Insta360 invaded DJI’s turf with the first mass-market panoramic drone, the Antigravity A1. It had instant appeal for drone pilots, offering 8K 360 video and features like subject tracking, obstacle detection and FPV flying.
Though beaten to the punch, DJI has responded quickly with the Avata 360, aided by its drone experience and camera tech from the new Osmo 360 action cam. It has a lot in common with its rival, but it’s safer to fly around people and offers single camera 4K footage on top of 8K 360 video.
To find out how the Avata 360 stacks up against the A1, I tested it both indoors and out, around people and even “stunt” horses. The Avata 360 isn’t perfect, but it is far more polished than its rival. As with other DJI drones of late, though, US availability remains unclear.
How the Avata 360 works
The 360 camera makes the Avata 360 different from any other DJI drone. It features two ultrawide cameras with f/1.9 lenses and 1.1-inch 64-megapixel sensors, with one pointing up and the other down for unobstructed 200-degree views. Those are then stitched together by software to create 360 degree video at up to 8K 60 fps.
This setup fundamentally changes the way you pilot a drone and capture video. Since the 360 camera records everything around it, you can focus on flying and reframe shots later in DJI’s Studio app. That 360 view is also handy when piloting in FPV mode. With the head tracking on DJI’s Goggles N3, you can look all around you simply by turning your head.
Unlike the Antigravity A1 that only shoots 360 video, the Avata 360 supports regular single-camera shooting as well. When you switch to that mode, the camera rotates forward and shoots 4K video at up to 60 fps with a 28mm field of view. DJI’s drone also works with a regular controller, which isn’t an option on the A1. You have to use the Insta360’s goggles at all times.
Design
Because of the chunky 360 camera, the Avata 360 is slightly bigger than the Avata 2. It’s also quite a bit heavier at 455 grams (one pound) and so, unlike the 249-gram Antigravity A1, you’ll need a permit to fly one in most regions.
The Avata 360’s propellers are shielded to protect the drone and keep it safe indoors or around people, unlike the open-prop A1. For additional protection it has two omni obstacle sensors on the side, a Lidar sensor up front, landing sensors on the bottom and, of course, a camera that points in all directions. To keep the lenses off of bare ground, it comes with a foldable 18x18-inch landing mat.
DJI’s Goggles N3 (available with the RC Motion 3 controller as an option) are comfortable and allow you to wear eyeglasses. Unlike the Goggles 3, though, there’s no external camera to see outside. If you’d rather pilot conventionally, you can get the Avata 360 bundled with DJI’s RC 2 screen controller in another kit.
Performance and features
As an FPV drone, the Avata 360 is fast and agile. It can hit speeds up to 40 MPH in sport mode (without obstacle avoidance) or 35 MPH in normal mode. When used with the optional FPV Remote Controller 3, you can do flips, rolls and other “cinewhoop” style maneuvers. Once you get used to flying it that way, it’s incredibly fun.
The 38.7Wh batteries have 26 percent more capacity than the Avata 2’s cells, but rated endurance is about the same at 24 minutes. I never got more than about 18 minutes in real-world flying though, so it’s a good idea to buy the Fly More kit with three batteries and a fast charger that can replenish them all in about 100 minutes.
Samuel Dejours for Engadget
Part of my testing of the Avata 360 was capturing “trick riding” horses, so it was important that the drone noise didn’t spook them. At 81db the Avata 360 is louder than the Mini 4 Pro (67 db) and less banshee-like than the Neo 2 due to the larger propellers, so the horses weren’t alarmed.
Video is transmitted to the Goggles N3 and RC Motion 3 controller (or the RC screen controller) at 1080p 60 fps via DJI’s OcuSync 4.0+ system. The maximum flying range is 20 km (12.4 miles), double that of the Avata 2. This is an impressive distance for an FPV drone.
The 45GB of internal storage (42GB usable) can fill up quickly when you’re shooting 8K video, but the drone also has a microSD slot. To get your footage onto a PC or smartphone, you can transfer it via the USB-C port or over Wi-Fi using DJI’s Fly app.
The Avata 360 has DJI’s usual tracking and obstacle detection features, but they’re available only with the RC 2 controller and not the Goggles N3. To follow a subject, simply draw a box around them on the controller to enable Focus Track and its three modes: Spotlight, Point of Interest and ActiveTrack. The latter tracks a subject automatically and lets you control the drone’s position via an on-screen “steering wheel.”
Steve Dent for Engadget
To test that, I biked in a narrow forested lane and walked around a bamboo-covered obstacle course. When using Focus Track, the drone dodged most obstacles and was only confused by small leaves and branches. If it did contact one of those, it sailed right through without crashing thanks to the propeller guards.
Subject tracking works in both 360 and single camera modes, but when using the latter, The Avata 360 can’t see and avoid obstacles behind it. It contacted branches several times during my testing, but fortunately the prop guards prevented crashes. Other automatic features include Dronie, Rocket and Quickshot modes that let you capture clips for social media.
Video
The Avata 360 prioritizes FPV freedom and flexibility over pure video quality. Though the specs promise 8K, that only applies to the full 360 degree video — your final, flat video will actually be 4K or less after processing. And the 360 camera zooms digitally (not optically), which further reduces resolution. On top of that, you can often see a “seam” in the video where stitching occurs, and dewarping (used to output flat video) can create softness at the edges. With all that, video is less sharp than DJI’s regular Mini, Air and Mavic drones.
Because the camera is fixed when shooting 360 video, the Avata 360’s gimbal can’t smooth out jolts or correct for roll. Instead, it uses action cam-style electronic stabilization. What’s more, that type of smoothing causes motion blur and artifacts in low light due to the lower shutter speeds — something I also noticed with the Osmo 360.
Steve Dent for Engadget
With that said, video quality was as sharp and color-accurate as the Osmo 360, and a touch better than the Antigravity A1 when filming in daylight. For tricky, contrasty conditions like a shaded path on a sunny day, the D-LogM option boosted dynamic range, helping me bring out shadow detail and tone down overly-bright highlights. However, the cityscape I shot at night was soft and occasionally blurry due to the aforementioned stabilization issues.
In exchange for this lower video quality, the 360 camera provides incredible flexibility. For a vlog style tracking shot, for instance, I usually need to film twice to show the forward and backward directions. With the Avata 360, though, I was able to get both POVs from the same shot and output an overhead view too for good measure.
DJI Studio is where you go to select your desired framing and output the flat video. It’s not quite as versatile as Insta360’s app, but it allows you to reframe shots and create smooth transitions between camera angles. And to save time, you can use the Intelligent Tracking feature to center your subject. The app also offers color correction and other features, but I found it easier to export the final shots to DaVinci Resolve for any additional work.
Flat video a mixed bag. Sharpness is mildly better than the 360 video, but the lack of a gimbal roll axis means that the camera can’t level itself when the drone banks into the wind. As a result, most of my footage was tilted and often unusable.
Wrap-up
Steve Dent for Engadget
With the Avata 360, DJI has a surprisingly sophisticated drone that offers better video quality and more features than its only rival, Insta360’s Antigravity A1. It can do everything you’d expect from a DJI FPV drone like tracking, obstacle avoidance and acrobatics, while operating safely around people (or horses).
DJI also managed to undercut Insta360 on price, with the Avata 360 starting at 459 euros (approx. $530), or 939 euros ($1,089) in a kit with a screen controller, three batteries and a charger. It’s also available with the extra batteries and charger, DJI’s FPV Goggles N3 and the RC Motion 3 controller for 939 euros ($1,089). As mentioned, there’s no word yet on US pricing, but the Avata 360 has been approved by the FCC so it could appear in the US soon.
Like 360 action cams though, this is a niche product. If you’re a vlogger, extreme action shooter, FPV pilot or solo filmmaker who wants speed and flexibility, the Avata 360 is a great choice. If it’s pixel perfect image quality you want, however, you’re better off with DJI’s Mini 5 Pro, Air 3S or Avata 4 Pro drone.
This article originally appeared on Engadget at https://www.engadget.com/cameras/dji-avata-360-drone-review-360-video-hits-new-highs-120014666.html?src=rss
OpenAI has "indefinitely" abandoned plans to release an a erotic chatbot for adults following concerns from employees and investors, the company confirmed to The Financial Times. Plans for such a feature, first announced in October 2025 for release in December last year, had already been delayed while company debated whether to release it all. It's the second app OpenAI has decided to shelf this week, after announcing on Tuesday that it was shutting down its Sora video generator.
The adult-oriented chatbot, reportedly called "Citron mode," is now on hold with no planned release date. The company reportedly had difficulty training models that previously avoided erotic content and also removing illegal behavior like bestiality or incest, two people familiar with the matter told the FT.
Open AI said that it wanted to conduct long-term research on the effects of erotic chats and user attachment to AI, adding that there was currently not yet enough "empirical evidence" on the subject. The company also said it wanted to focus on its core productivity tools like coding assistants and drop "side quests" like Sora and the erotic chatbot.
The idea for adult features came after OpenAI announced that it would add parental controls and automatic age detection features for ChatGPT. CEO Sam Altman said back in October that the company had always been careful about such issues over concerns around unhealthy AI attachments, but felt comfortable that it could "safely relax the restrictions in most cases."
However, the adult mode had reportedly caused concern among investors, particularly amid the controversy caused by rival xAI's Grok model that generated deepfake nudes of real people and children. Staff also worried about the feature, with one senior employee even leaving the company over the issue. "AI shouldn’t replace your friends or your family; you should have human connections," he told the FT.
Another challenge is OpenAI's age-checking tech, introduced following lawsuits from families who said that ChatGPT harmed their children. The tech reportedly has an error right higher than ten percent, which would still give a large number of young people access to the tech. OpenAI said that figure is in the industry standard range and that it is continuing to work on its accuracy.
This article originally appeared on Engadget at https://www.engadget.com/ai/openai-drops-plans-to-release-an-adult-chatbot-113121190.html?src=rss