News
Elon Musk shares Tesla FSD Beta 9.2’s behind-the-scenes technical release notes
Tesla released FSD Beta 9.2 over the weekend, and while the update’s official Release Notes were identical to FSD Beta 9.2, there were a lot of improvements that were rolled out by the company under the hood. These improvements were so notable that Elon Musk stated on Twitter that 9.2’s release had to be moved back a bit to address some last-minute issues.
The Tesla CEO then decided to share the technical release notes for FSD Beta 9.2, which explains every optimization that has been introduced in the recent update. Based on what Musk has shared, it appears that FSD Beta 9.2 is a massive step up from V9.1, allowing the advanced driver-assist system to navigate through inner-city streets in a safer and more confident manner.
The following are the technical release notes for FSD Beta 9.2, as shared by CEO Elon Musk on Twitter.
- Clear-to-go boost through turns on minor-to-major roads (plan to expand to all roads in V9.3).
- Improved peek behavior where we are smarter about when to go around the lead vehicle by reasoning about the causes for lead vehicle being slow.
- V1 of the Multi-modal prediction for where other vehicles expected to drive. This is only partially consumed for now.
- New Lanes network with 50k more clips (almost double) from the new auto-labeling pipeline.
- New VRU velocity model with 12% improvement to velocity and better VRU clear-to-go performance. This is the first model trained with “Quantization-Aware-Training”, an improved technique to mitigate int8 quantization.
- Enabled Inter-SoC synchronous compute scheduling between vision and vector space processes. Planner in the loop is happening in v10.
- Shadow mode for new crossing/merging targets network which will help improve VRU control.
Apart from the technical details of FSD Beta 9.2’s release, Elon Musk also hinted at the upcoming updates coming for the advanced driver-assist system. The CEO noted that FSD Beta 9.2 would be followed by V9.3 and V9.4, but after this point, Tesla may already release FSD Beta 10. V10 would feature “significant architecture changes,” as per the CEO. Musk did not explain further, though speculations suggest that FSD Beta 10 would be the update that would transition Tesla into a single FSD stack.
Lastly, Musk provided some updates on the release date of the FSD Beta Download Button, which has experienced substantial delays. The CEO explained that the “Button” would probably roll out with Beta 10 or V10.1, which, based on the current release cycles of FSD Beta’s updates, would likely be in the next couple of months, barring any unexpected delays.
“Beta 10 or maybe 10.1. Going to pure vision set us back initially. Vision plus (coarse) radar had us trapped in a local maximum, like a level cap. Pure vision requires fairly advanced real-world AI, but that’s how our whole road system is designed to work: NN’s with vision,” Musk explained.
Don’t hesitate to contact us with news tips. Just send a message to tips@teslarati.com to give us a heads up.
Elon Musk
Tesla confirmed HW3 can’t do Unsupervised FSD but there’s more to the story
Tesla confirmed HW3 vehicles cannot run unsupervised FSD, replacing its free upgrade promise with a discounted trade-in.
Tesla has officially confirmed that early vehicles with its Autopilot Hardware 3 (HW3) will not be capable of unsupervised Full Self-Driving, while extending a path forward for legacy owners through a discounted trade-in program. The announcement came by way of Elon Musk in today’s Tesla Q1 2026 earnings call.
🚨 Our LIVE updates on the Tesla Earnings Call will take place here in a thread 🧵
Follow along below: pic.twitter.com/hzJeBitzJU
— TESLARATI (@Teslarati) April 22, 2026
The history here matters. HW3 launched in April 2019, and Tesla sold Full Self-Driving packages to owners on the understanding that the hardware was sufficient for full autonomy. Some owners paid between $8,000 and $15,000 for FSD during that period. For years, as FSD’s AI models grew more demanding, HW3 vehicles fell progressively further behind, eventually landing on FSD v12.6 in January 2025 while AI4 vehicles moved to v13 and then v14. When Musk acknowledged in January 2025 that HW3 simply could not reach unsupervised operation, and alluded to a difficult hardware retrofit.
The near-term offering is more concrete. Tesla’s head of Autopilot Ashok Elluswamy confirmed on today’s call that a V14-lite will be coming to HW3 vehicles in late June, bringing all the V14 features currently running on AI4 hardware. That is a meaningful software update for owners who have been frozen at v12.6 for over a year, and it represents genuine effort to keep older hardware relevant. Unsupervised FSD for vehicles is now targeted for Q4 2026 at the earliest, with Musk describing it as a gradual, geography-limited rollout.
For HW3 owners, the over-the-air V14-lite update is welcomed, and the discounted trade-in path at least acknowledges an old obligation. What happens next with the trade-in pricing will define how this chapter ultimately gets written. If Tesla prices the hardware path fairly, acknowledges what early adopters are owed, and delivers V14-lite on the June timeline it committed to today, it has a real opportunity to convert one of the longest-running sore subjects among early adopters into a loyalty story.
Elon Musk
Tesla isn’t joking about building Optimus at an industrial scale: Here we go
Tesla’s Optimus factory in Texas targets 10 million robots yearly, with 5.2 million square feet under construction.
Tesla’s Q1 2026 Update Letter, released today, confirms that first generation Optimus production lines are now well underway at its Fremont, California factory, with a pilot line targeting one million robots per year to start. Of bigger note is a shared aerial image of a large piece of land adjacent to Gigafactory Texas, that Tesla has prominently labeled “Optimus factory site preparation.”
Permit documents show Tesla is seeking to add over 5.2 million square feet of new building space to the Giga Texas North Campus by the end of 2026, at an estimated construction investment of $5 billion to $10 billion. The longer term production target for that facility is 10 million Optimus units per year. Giga Texas already sits on 2,500 acres with over 10 million square feet of existing factory floor, and the North Campus expansion is being built to support multiple projects, including the dedicated Optimus factory, the Terafab chip fabrication facility (a joint Tesla/SpaceX/xAI venture), a Cybercab test track, road infrastructure, and supporting facilities.
Credit: TESLA
Texas makes strategic sense beyond the existing infrastructure. The state’s tax structure, lower labor costs relative to California, and the proximity to Tesla’s AI training cluster Cortex 1 and 2, both located at Giga Texas and now totaling over 230,000 H100 equivalent GPUs, means the Optimus software stack and the factory producing the hardware will share the same campus. Tesla’s Q1 report also confirmed completion of the AI5 chip tape out in April, the inference processor designed specifically to power Optimus units in the field.
As Teslarati reported, the Texas facility is intended to house Optimus V4 production at full scale. Musk told the World Economic Forum in January that Tesla plans to sell Optimus to the public by end of 2027 at a price between $20,000 and $30,000, stating, “I think everyone on earth is going to have one and want one.” He has previously pegged long term demand for general purpose humanoid robots at over 20 billion units globally, citing both consumer and industrial use cases.
Tesla (NASDAQ: TSLA) reported its earnings for the first quarter of 2026 on Wednesday afternoon. Here’s what the company reported compared to what Wall Street analysts expected.
The earnings results come after Tesla reported a miss on vehicle deliveries for the first quarter, delivering 358,023 vehicles and building 408,386 cars during the three-month span.
As Tesla transitions more toward AI and sees itself as less of a car company, expectations for deliveries will begin to become less of a central point in the consensus of how the quarter is perceived.
Nevertheless, Tesla is leaning on its strong foundation as a car company to carry forward its AI ambitions. The first quarter is a good ground layer for the rest of the year.
Tesla Q1 2026 Earnings Results
Tesla’s Earnings Results are as follows:
- Non-GAAP EPS – $0.41 Reported vs. $0.36 Expected
- Revenues – $22.387 billion vs. $22.35 billion Expected
- Free Cash Flow – $1.444 billion
- Profit – $4.72 billion
Tesla beat analyst expectations, so it will be interesting to see how the stock responds. IN the past, we’ve seen Tesla beat analyst expectations considerably, followed by a sharp drop in stock price.
On the same token, we’ve seen Tesla miss and the stock price go up the following trading session.
Tesla will hold its Q1 2026 Earnings Call in about 90 minutes at 5:30 p.m. on the East Coast. Remarks will be made by CEO Elon Musk and other executives, who will shed some light on the investor questions that we covered earlier this week.
You can stream it below. Additionally, we will be doing our Live Blog on X and Facebook.
Q1 2026 Earnings Call at 4:30pm CT https://t.co/pkYIaGJ32y
— Tesla (@Tesla) April 22, 2026
Tesla confirmed HW3 can’t do Unsupervised FSD but there’s more to the story
Tesla isn’t joking about building Optimus at an industrial scale: Here we go
