• Latest
  • Trending
  • All
Answer card summarising that Tesla is installing Optimus production lines at Fremont for 2026 while the V3 robot itself remains unrevealed, with only one hardware figure ever published.

Tesla Optimus V3: production starts, reveal still missing

28 July 2026
Answer card stating that OpenAI released the Agents API in public beta on 10 September 2026 with no separate fee, billed through model tokens, tool calls and hosted sandbox time, with a choice of OpenAI hosted, self hosted or partner sandboxes, US only data residency and no Zero Data Retention support.

OpenAI’s Agents API has no fee, no ZDR and a one hour sandbox clock

14 September 2026
Answer card: Sakana Fugu Max at $2 and $6 per million tokens, Fugu Ultra v2 unchanged at $5 and $30, and Sakana saying Ultra v2 scores without Fable 5 or GPT-6 Astra in its pool.

Fugu Max costs $2 and $6 while Fugu Ultra v2 runs without Fable 5

13 September 2026
Answer card stating that DeepSeek released DeepSeek-V4.1-Flash on 10 September 2026 as a 552 billion parameter mixture of experts model with a new causal encoder decoder architecture that activates 8 billion parameters on input and 16 billion on output, with native vision, a one million token context and MIT licensed weights, that the API model name is now deepseek-flash at 0.15 dollars per million input tokens and 0.60 dollars per million output tokens off peak, and that DeepSeek announced V4 Pro would be routed to V4.1-Flash from 14 September and reversed that on 11 September.

DeepSeek V4.1-Flash arrived, and the V4 Pro retirement lasted a day

12 September 2026
Answer card stating that Cognition released SWE-2 on 10 September 2026, a coding model post-trained from Kimi K3, scoring 50.0 percent on FrontierCode 1.1 Main against 50.9 percent for Claude Fable 5.1 and 27.3 percent on Terminal-Bench 4 against 55.8 percent, available only inside Devin.

SWE-2 trails Fable 5.1 by one point, and by 28 on Terminal-Bench 4

11 September 2026
Answer card for Meta Muse, free to 100 million tokens a week then $20 a month, launched 8 September 2026 for United States adults only, running in a dedicated per user virtual machine.

Does Meta Muse do enough to earn your inbox and a card on file?

9 September 2026
Answer card stating that the public download pages for the VMware Virtual Disk Development Kit on developer.broadcom.com began returning 404 errors on 25 August 2026 with no announcement or deprecation notice, that Broadcom support tells customers the kit is no longer available for use or download, and that release lines 7.0.3.1, 8.x and 9.x are all affected.

Broadcom pulled VDDK 8.0 and 9.0, and the 404 is the only notice

8 September 2026
Answer card stating that OpenAI published its research acceleration measurements on 6 September 2026, that as of mid August 2026 its research organisation logged 3.1 agent workdays of coding agent runtime for every workday of human labour normalised to a standard eight hour day, and that OpenAI states this should not be read as a 3.1 times productivity gain because it measures runtime rather than delivered output.

OpenAI’s 3.1 agent-workdays per human day is not a 3.1x gain

7 September 2026
Answer card stating that Mullvad announced on 3 September 2026 that it is shutting down its public encrypted domain name system servers on 2 November 2026 and sponsoring the Quad9 Foundation instead, with 194.242.2.2 and its five sibling addresses all going away, and virtual private network customers unaffected.

Mullvad’s DNS servers go dark on 2 November, and Quad9 blocks no ads

5 September 2026
OpenAI announcement image for GPT-6 Astra, a spiral galaxy of white, blue and amber points of light curling around a bright core on a near black star field.

GPT-6 Astra lists at $10 and $50, 2.5x what GPT-5.6 Sol costs

6 September 2026
Google's official announcement image for the release, reading Introducing Gemini 3.8 Flash and 3.8 Flash Cyber in black type over a pale blue background with a blurred white chevron and the four colour Gemini spark below.

Gemini 3.8 Flash keeps the price and the 1 January cliff

3 September 2026
Answer card stating that Anthropic announced Enterprise Frontier Safeguards on 1 September 2026, that activity data used for misuse monitoring moves into cloud storage the customer controls under the customer own encryption keys, that Anthropic charges nothing for the feature while the cloud provider bills storage and egress, and that the phased rollout starts later in autumn 2026 with interim zero data retention on Fable 5 and Fable 5.1 for eligible customers.

Anthropic moves retention into your own cloud, for 30 days

3 September 2026
Official Google diagram of a client connection in three numbered steps: a DNS lookup with a query and an address, a TLS ClientHello and ServerHello, then a content exchange with a website. A callout on the DNS step reads 25% of global web traffic is now protected by encrypted DNS, and a callout beside an Android phone on the ClientHello step reads Android 17 supports ECH GREASE by default.

Android 17 hides the SNI, not your DNS or destination

3 September 2026
  • About
  • Contact
  • Privacy
  • Legal
Tuesday, September 15, 2026
  • Login
Packet Nebula
  • Home
  • Articles
    • Security
    • Network
    • Dev
    • Sysadmin
    • SEO
    • Email & DNS
  • Tools
    • Network tools: free, fast, no signup
    • Security tools: free, fast, no signup
    • Developer tools: free, fast, no signup
    • Sysadmin tools: free, fast, no signup
    • SEO tools: free, fast, no signup
    • Email & DNS tools: free, fast, no signup
  • Download
  • About
No Result
View All Result
Packet Nebula
No Result
View All Result
Home Dev

Tesla Optimus V3: production starts, reveal still missing

by stephane
28 July 2026
in Dev
0
Answer card summarising that Tesla is installing Optimus production lines at Fremont for 2026 while the V3 robot itself remains unrevealed, with only one hardware figure ever published.
493
SHARES
1.4k
VIEWS
Share on FacebookShare on Twitter

Picture a factory floor in Fremont where the last Model S rolled off in early May, now filling up with tooling for a machine nobody outside Tesla has laid eyes on. That's the Optimus V3 situation as of this week: the production lines are going in, and the robot has still never been revealed. Tesla told shareholders on 22 July that first-generation Optimus lines are being installed, with production anticipated in 2026. The same call set no unveiling date, gave no unit target, and disclosed no dimensions. Musk spent most of his Optimus minutes lowering expectations instead, calling it the hardest product Tesla has ever tried to scale, with no existing supply chain behind it. So the spec sheets you're finding in search results? Somebody made those up.

The short answer

Tesla is installing Optimus production lines at Fremont for a robot it has never shown. The Q2 call on 22 July confirmed the tooling and set no reveal date, no unit target and no specs. First units go to internal training data collection, not customers. Every detailed V3 spec sheet online recycles Gen 2 numbers or invents them.

0public sightings of V3
1hardware figure Tesla published
10,000unique parts, per Musk
Answer card: Tesla is installing first-generation Optimus production lines at Fremont for production in 2026, the V3 reveal has slipped from Q1 to mid-year to no date at all, Musk calls it the hardest product Tesla has scaled with no existing supply chain and about 10,000 unique parts, and only one hardware number has ever been published.
The whole story in one card. A factory schedule moving forward, attached to a machine nobody has seen.

What Tesla actually said

The Q2 2026 call ran on 22 July, and the Optimus section was shorter on news than anyone expected. Tesla’s own line: first-generation production lines for Optimus are being installed, in anticipation of production in 2026. That’s the concrete part. It sits on the Fremont floor that built Model S and X until early May, which is a real conversion of a real building, not a slide.

Everything after that was Musk managing expectations downward. He called it the hardest product to scale in manufacturing Tesla has ever made, because everything on the robot is new. He said there’s no existing supply chain for it. Back in April he’d put it more bluntly: the rate moves as fast as the least lucky, slowest, dumbest part in the entire ten thousand. And he declined to name a 2026 unit target, which is a notable retreat from January 2025, when he floated roughly 10,000 robots for that year and Tesla built none.

The forward-looking bits were about generations that don’t exist yet. Optimus 4 will be more vertically integrated, he said. AI5 will go into Optimus eventually. Fine, but that’s roadmap talk about a successor to a robot we haven’t met.

Timeline showing the Optimus V3 reveal date moving from a promised Q1 2026 in October 2025, to middle of this year on the 22 April 2026 call, to no date set on the 22 July Q2 call, while production lines were installed anyway.
Three public updates, three different answers, and the factory schedule kept moving regardless.

The reveal that keeps not happening

Musk originally said Tesla would unveil V3 “probably in Q1” 2026. April came and the answer became “probably middle of this year.” July came, and the Q2 call simply didn’t address it.

Tesla has offered a reason, and honestly it’s the most credible part of the whole story. The company says rivals run frame-by-frame analysis whenever it releases footage and copy whatever they can lift, so the design stays under wraps until production is actually running. Anyone who has watched the humanoid field this year knows the concern isn’t imaginary. Designs converge fast.

But secrecy has a cost that Tesla doesn’t pay and you do. Nobody outside the company can assess this machine. Not its reach, not its payload, not its duty cycle, not what it costs to keep running for a shift. Those are the numbers that decide whether a humanoid is a tool or a demo, and for V3 they’re all blank.

Every V3 spec sheet is made up

This is the part worth being loud about, because the search results are genuinely polluted.

Tesla has published exactly one hardware number for this robot: 50 actuators per unit, 25 in each forearm and hand assembly. That’s it. And Musk has since said the mechanical design moved on from that description, so even the one number is stale. Add “roughly 10,000 unique parts” if you’re feeling generous about what counts as a spec.

Now look at what’s circulating. A height of 173 cm and a weight of 57 kg, quoted confidently across dozens of pages. Those are Gen 2 figures, relabelled. A $30,000 price, which Tesla has never announced for any version. An IP68 ingress rating, a 20 km/h walking speed, 22 degrees of freedom per hand. No Tesla source exists for any of them. And a persistent claim that mass production began back in January, which contradicts Tesla’s own July statement that the lines are still being installed.

I find this more interesting than annoying, actually. It’s a clean example of what happens when search demand for a number vastly exceeds the supply of the number. The gap fills itself, the filler gets scraped, the scrape gets cited, and eighteen months later a fabricated height is treated as a fact. We ran into the same pattern with Etched’s Sohu projections, where 2024 figures for a discontinued design still get quoted as current.

Checklist splitting confirmed Optimus V3 facts from invented ones: the Fremont line installation, the 50 actuators figure, the 10,000 parts and the internal-use first units are on the record, while the 173 cm and 57 kg dimensions, the 30,000 dollar price, the IP68 rating and the claim that mass production already began are not from Tesla.
The line between what Tesla has said and what got written on its behalf.

What it changes for you

If you’re evaluating humanoids for actual work this year, the practical answer is that Optimus isn’t evaluable. You can’t run a comparison against a robot with no published payload. Tesla itself said the first units go to training data collection and further functionality development, which is internal language for “we’re still learning what it can do.”

Compare that with what a shippable pitch looks like. Humanoid’s HMND 01 Alpha publishes its height, its degrees of freedom, its payload and the plants where it ran pilots, and you can argue with those numbers precisely because they exist. Same on the software side, where Mistral’s Robostral Navigate put a benchmark score on the table for people to poke at. Disclosure invites criticism. That’s the trade Tesla has decided not to make.

Timing note, since it’s this week: IEEE Humanoids 2026 runs 28 to 31 July in Santa Clara, with “Humanoids and the Future of Work” as the special topic for its 25th edition. Tesla’s factory sits a short drive north. One of those two things will produce published, reviewable numbers about humanoid labour, and it won’t be the factory.

My read, and I might be wrong here: the Fremont conversion is the real signal and the missing reveal is noise. Converting an S and X line is expensive and hard to reverse, so somebody at Tesla believes in the schedule. The silence probably says more about competitive paranoia than about the robot being in trouble. Just don’t let anyone sell you a spec sheet in the meantime.

Sources: the production line installation, the scaling remarks and the Optimus 4 and AI5 comments come from Tesla’s Q2 2026 earnings call on 22 July 2026. The reveal slipping to “middle of this year”, the late July or August production start, the roughly 10,000 unique components and the frame-by-frame copying rationale come from Electrek’s report on the Fremont Model S and X line conversion (22 April 2026). The audit of circulating V3 specifications, including the recycled Gen 2 dimensions and the unsourced price, is documented in RoboZaps’ Optimus Gen 3 status page. Conference dates are from the IEEE Robotics and Automation Society listing for Humanoids 2026. No Optimus V3 dimensions, price or performance figures are asserted here, because Tesla has published none.

Frequently asked questions

Has Tesla revealed Optimus V3?

No. As of late July 2026 the third-generation Optimus has not been shown publicly. Musk first said the unveiling would come "probably in Q1" 2026, then pushed it to "probably middle of this year" on the April earnings call, and the 22 July Q2 call set no new date. Tesla has said it is holding the design back because competitors analyse its robot videos frame by frame, so it plans to keep V3 hidden until production is under way.

When does Optimus production actually start?

Tesla told shareholders on 22 July 2026 that first-generation Optimus production lines are being installed at Fremont, with production anticipated during 2026. On the April call Musk pointed at late July or August for the start, roughly four months after the last Model S and X came off that line in early May. He also warned output would be slow and gave no unit target for the year, so treat "production starts" as a line switching on rather than robots shipping.

What are the confirmed Optimus V3 specs?

Almost nothing. The one hardware figure Tesla has stated publicly is 50 actuators per robot, 25 in each forearm and hand assembly, and Musk has since said the mechanical design changed after that. Tesla has also described roughly 10,000 unique parts. No official height, weight, payload, battery, speed or ingress rating exists for V3. Anything more detailed than that is coming from somewhere other than Tesla.

Is the $30,000 Optimus price real?

It is not a Tesla number. No price has been announced for Optimus V3, there is no order page and no configurator. Musk has talked in the past about eventual low unit costs at volume, which is a manufacturing ambition rather than a price. The first robots off the Fremont line are earmarked for internal training data collection and functionality development, not for sale, so a consumer price would be premature anyway.

Why do so many sites list detailed Optimus V3 specifications?

Because the demand for those numbers exists and the supply does not, so the gap gets filled. The commonly quoted 173 cm and 57 kg figures are Gen 2 numbers reused under a V3 heading. Other details making the rounds, an IP68 rating, a 20 km/h walking speed, 22 degrees of freedom per hand, have no Tesla source behind them. If a spec sheet does not cite a Tesla filing, video or executive statement, assume it was assembled from older material and guesswork.

Tags: hardwarehumanoidmanufacturingnewsroboticstesla
Share197Tweet123
stephane

stephane

  • Trending
  • Comments
  • Latest
Answer card: Proton Lumo 2.0 is private by policy, not by locality. Saved history is locked so even Proton cannot read it, but the prompt is decrypted on a Proton EU server to answer it, then forgotten.

Proton Lumo 2.0 review: how private is it, really?

3 September 2026
The Agentic Coding section of the official Hy4 preview benchmark appendix published by Tencent, a table comparing Hy3 and Hy4 preview against DeepSeek V4 Pro 0813, Qwen 3.8 Max, GLM 5.3, Kimi K3, GPT 5.6 Sol and Claude Opus 5 across SWE-bench Multilingual, SWE-bench Pro, DeepSWE, three SWE Atlas tasks, SWE-Marathon, Terminal-Bench 2.1, NL2Repo-Bench, CyberGym, ProgramBench, PostTrainBench and Harbor-Index.

Tencent’s 770B Hy4 tops one benchmark row in 46

3 September 2026
Answer card: Qwen 3.7 Max is API-only and cannot run locally yet; the open Qwen models (Qwen 3.6 27B, qwen3:8b to 32b) run offline via Ollama.

Qwen 3.7 local: what you can actually run offline

22 June 2026
Answer card: JWTs are not encrypted, anyone can read them; the signature proves who issued the token, not who may read it.

Are JWTs encrypted? No, and the difference will bite you

0
Answer card: a random 8 character password falls in under 2 hours offline, while 16 random characters hold for 1.4 trillion years at the same speed.

How long does it take to crack a password in 2026?

0
Answer card: three DNS records decide if your mail lands or bounces; SPF lists allowed senders, DKIM signs messages, DMARC sets the failure policy.

SPF, DKIM and DMARC explained: the records your email needs

0
Answer card stating that OpenAI released the Agents API in public beta on 10 September 2026 with no separate fee, billed through model tokens, tool calls and hosted sandbox time, with a choice of OpenAI hosted, self hosted or partner sandboxes, US only data residency and no Zero Data Retention support.

OpenAI’s Agents API has no fee, no ZDR and a one hour sandbox clock

14 September 2026
Answer card: Sakana Fugu Max at $2 and $6 per million tokens, Fugu Ultra v2 unchanged at $5 and $30, and Sakana saying Ultra v2 scores without Fable 5 or GPT-6 Astra in its pool.

Fugu Max costs $2 and $6 while Fugu Ultra v2 runs without Fable 5

13 September 2026
Answer card stating that DeepSeek released DeepSeek-V4.1-Flash on 10 September 2026 as a 552 billion parameter mixture of experts model with a new causal encoder decoder architecture that activates 8 billion parameters on input and 16 billion on output, with native vision, a one million token context and MIT licensed weights, that the API model name is now deepseek-flash at 0.15 dollars per million input tokens and 0.60 dollars per million output tokens off peak, and that DeepSeek announced V4 Pro would be routed to V4.1-Flash from 14 September and reversed that on 11 September.

DeepSeek V4.1-Flash arrived, and the V4 Pro retirement lasted a day

12 September 2026
  • About
  • Contact
  • Privacy
  • Legal

Copyright © 2026 Stephane Cardon.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • Home
  • Articles
    • Security
    • Network
    • Dev
    • Sysadmin
    • SEO
    • Email & DNS
  • Tools
    • Network tools: free, fast, no signup
    • Security tools: free, fast, no signup
    • Developer tools: free, fast, no signup
    • Sysadmin tools: free, fast, no signup
    • SEO tools: free, fast, no signup
    • Email & DNS tools: free, fast, no signup
  • Download
  • About

Copyright © 2026 Stephane Cardon.