CLARITY Act: 60 Votes Left to Pass, Market Puts Odds at Only 17%

marsbitPublished on 2026-09-11Last updated on 2026-09-11

Abstract

The CLARITY Act (H.R. 3633), which aims to definitively assign regulatory jurisdiction over digital assets to either the SEC or CFTC, faces a critical procedural vote in the US Senate on September 15th. This cloture vote requires 60 "yes" votes to allow debate to proceed. With 53 Republican seats, at least 7 Democratic votes are needed, but key Democrats remain opposed, primarily due to the bill's ethics provisions. Some Republican senators also may vote "no." Failure to reach 60 votes would likely kill the bill for 2026 due to the truncated legislative calendar. Prediction markets like Polymarket estimate only a 17.5% chance of the bill becoming law this year, down from 82% in February. Significant money is betting against its passage. If the cloture vote passes, the bill would establish a lasting regulatory framework: classifying tokens as securities or commodities, requiring CFTC registration for intermediaries, providing legal safe harbors for developers and self-custody, and implementing soft ethics rules for officials. If it fails, regulatory clarity would shift to agencies like the SEC and CFTC through rulemaking, which is seen as less permanent and more susceptible to change. The industry is divided, with some urging negotiation to save the bill and others prepared to accept a fragmented, issue-by-issue legislative approach if it fails.

At 2:15 PM Eastern Time on September 15, the US Senate will vote on a bill.

This vote does not decide whether it becomes law; it only determines whether senators can begin formal debate on it. But it is likely the endpoint for this bill.

To clear this hurdle, 60 votes in favor are needed. The Republican party holds 53 seats.

The bill's full name is the Digital Asset Market Clarity Act, numbered H.R. 3633. Its goal is singular: to clarify whether a token falls under the SEC's jurisdiction or the CFTC's.

This debate has raged for a decade. In July 2025, the House of Representatives passed it once with a 294-134 vote, with 78 Democrats voting in favor. Then it stalled in the Senate for 14 months.

On September 10, Republicans Released a New Text

On Thursday, Republican Senator Cynthia Lummis unveiled the updated full text, approximately 630 pages long. She stated this text "reflects the hard, bipartisan work during the August recess," incorporating over 100 changes requested by Democrats (some reports say 114 to 115).

The changes, however, are only in three areas.

First, "nominally decentralized" protocols must register. If a trading protocol is not truly decentralized—meaning a person or group, through contract, arrangement, or relationship, has the authority to directly or indirectly control or materially alter its functions, operations, or consensus rules—it must register with the CFTC and comply with Bank Secrecy Act obligations. A carve-out exists: participating in governance or security committees does not by itself constitute "control"; a protocol must be one that can be manipulated by a specific entity. Subsequent CFTC rulemaking will focus on spot and cash digital commodity markets.

Second, DeFi's scope of application has been narrowed. The related protections only cover spot and cash transactions of digital commodities, explicitly excluding prediction markets. This aims to address objections from Native American tribes with gaming interests.

Third, credit unions' authority is clarified. Federal credit unions may use digital assets or distributed ledger systems to conduct any business they are legally permitted to do.

The unchanged parts are the crucial ones.

The ethics provision remains word-for-word unchanged. It's still the version approved by the White House in July: high-level officials including the President, Vice President, members of Congress, and their spouses are prohibited from issuing or sponsoring digital assets until January 20, 2029, with no retroactive effect, and enforcement lies solely with the Department of Justice—state attorneys general and private lawsuits have no enforcement power.

The BRCA (protecting non-custodial software developers from being treated as money transmitters) and stablecoin yield provisions also remain the July version.

How Long Has This Path Been Walked?

Those 7 Democrats are Alsobrooks, Booker, Cortez Masto, Gallego, Hickenlooper, Warner, and Warnock. Gallego said directly at the time that the ethics provision "is not a serious proposal."

Since the release of the new text, not one of these 7 has publicly shifted their stance.

What Is This Vote Really About?

Let's be clear first: This is not the final passage vote. The cloture motion is a procedural vote to "end debate"; passing it only means formal debate and amendment proposals can begin.

The math is simple. Republicans have 53 seats. If they all stay united, they are still short by 7 Democratic (or independent) senators.

The trouble is that Republicans aren't unified either. Media reports have named Rand Paul and Josh Hawley as opponents, citing substantive reasons; Tillis has stated he won't support it unless the ethics provision is strengthened. If these three all defect, the needed Democratic votes become 10 or more.

The Senate only reconvenes on September 14, with the vote the day after. That leaves less than 5 days for both sides.

Thus, the real meaning of this vote is: If it doesn't reach 60, this bill is essentially dead for 2026. The Senate will be out of session for most of October, with midterm elections on November 3rd. Some media have calculated: after this vote, the number of working days left in this session to complete the bill's process is countable on one hand.

What Are the Market Odds? Look at Polymarket First.

The main market asks only one question: Will the CLARITY Act (H.R.3633) be signed into law within 2026?

How that curve collapsed: It was at 82% in February, dropped to 65% before the August recess, fell to 17% after the vote was postponed, and remained at 17.5% as of September 10.

Now look at the sub-markets. A crucial caveat here: The market asking "how many senators will vote in favor" settles based on "the Senate's first final passage vote," not the September 15 vote—market rules explicitly exclude procedural motions, cloture, and amendment votes. In other words, these numbers measure "if this step passes, how far can it go afterwards?"

The 29% figure must be viewed alongside the 17.5%. The probability of reaching a final passage vote *and* securing over 60 votes is 29%, but the probability of actually becoming law within the year is only 17.5%. This gap corresponds to the remaining processes not yet completed: full debate, amendments, a second cloture, reconciliation with the House version, and presidential signature.

Another market bets on specific individuals. Again, note: the settlement basis is still "final passage vote," not the September 15 vote. So the numbers below are not "will he vote to advance?" but "will he vote in favor on final passage?"—if there is no final vote later, it's settled as No.

The most glaring are the last three rows. Tillis and Gallego are the co-negotiators on the ethics provision, yet the market gives them only 20% and 15.5% odds respectively. One demands the ethics provision be strengthened to support it, the other said in July this year that the ethics text "is not a serious proposal."

Here's also a discrepancy: Polymarket gives Rand Paul 38.5%, but media reports say Paul opposes the bill on substantive grounds. The market's pricing is inconsistent with the senator's public stance—a discrepancy worth noting in itself.

Money is talking too. On Polymarket, six identified wallets have collectively bet over $3.6 million on "No," including two newly created anonymous accounts placing approximately $818,000 and $677,000 respectively. These "No" positions were built just before the vote.

On the other side are only two cross-market veterans, TwoEyes and Geminae.Columbae, still holding Yes positions, with entry costs corresponding to about a 20% probability of passage, now sitting on a floating loss of about $11,800. But they simultaneously hold approximately $19.92 million in crypto longs on Hyperliquid—this looks more like buying political insurance for their massive long positions than betting on the bill itself.

Other estimates: Galaxy Research estimates "low double digits," around 10%.

And then there's a leak. Andy, host of The Rollup, said on X that sources from Capitol Hill put the real probability at only 3% to 5%, and insiders in Washington widely believe it won't pass, but no one can say so publicly because the industry has already poured tens of millions of dollars and 18 months into it. This is unconfirmed, but it's the most widely circulated "insider estimate" at the moment.

If It Passes, What Happens?

In one sentence: The industry gets enduring statutory law, not the whims of an administration.

Specifically, four things.

First, the division of labor between the SEC and CFTC is codified into law. Tokens on decentralized, mature blockchains (BTC, ETH meet the "mature blockchain" standard) have their spot markets regulated by the CFTC; "investment contract assets" still dependent on team effort or financing fall under the SEC, with a path provided for transitioning from SEC to CFTC. The decade-long debate over "Is a token a security or a commodity?" gets a statutory answer.

Second, intermediaries must follow rules. Digital commodity exchanges, brokers, and dealers must register with the CFTC, requiring customer asset segregation, qualified custody, disclosure, and market surveillance—an institutional response aimed at preventing another FTX-type collapse. The merged text from July also included a structural adjustment: changing registration from "provisional registration" to a "notice of intent" process, and adding new registration categories for digital commodity pool operators, digital commodity trading advisors, and their associated persons (this part came from the July 22 version, not in the September 10 revision).

Third, developers and self-custody get a legal floor. The BRCA provides a safe harbor for non-custodial software developers: they are not treated as money transmitters merely for writing code or maintaining a network. After the DOJ prosecuted Samourai Wallet developers, this provision's nature changed completely—it effectively adds a statutory shield against the enforcement logic of "publishing code is illegal." The new text also incorporates protections for self-custody wallets originating from the Keep Your Coins Act.

Fourth, the ethics provision takes effect, but it's weak. High-level officials and their spouses are barred from issuing or sponsoring digital assets until January 20, 2029, no retroactive effect, and enforcement is solely by the Department of Justice.

What happens at the market level? Tokens explicitly classified as "digital commodities" get a clarity premium; compliant exchanges gain a moat. But don't mistake cloture passage for smooth sailing—the developer safe harbor and self-custody protections are precisely the two items most likely to be used as bargaining chips during the amendment stage.

If It Fails, It's Not "Status Quo"

This is the most common misconception.

Failure does not mean staying put; it means legislative authority shifts from Congress to regulatory agencies.

The SEC has already taken a step. In August 2026, it proposed "Regulation Crypto Assets," consisting of four parts: a startup exemption (up to $5 million per offering, max 4 years), a financing exemption, a non-exclusive conditional safe harbor for "investment contracts" under the Howey test, and preemption of certain state registration and qualification requirements.

CFTC Chairman Michael Selig also spoke bluntly in a Bloomberg TV interview on August 20: "Establishing market structure is very important. We can do that either through regulatory rulemaking or through legislation." Legislation is the most reliable way to lock in market structure, but agencies also have significant authority under existing statutes and will use it when needed.

What's the difference? Statutes provide safe harbors and federal preemption; agency rules lead to comment letters, enforcement actions, and court summons. The former is more durable; the latter is easier to overturn by the next administration. Moreover, as some scholars point out, whether the severely understaffed CFTC has the capacity to write rules, register entities, build systems, and oversee markets is itself a question.

The legislative window will also effectively close. A more likely path is splitting the bill: no longer pursuing one comprehensive bill, but legislating separately on stablecoins, asset tokenization, perpetual contracts, prediction markets. According to Capitol Hill sources, this is a more realistic long-term path.

One thing won't stop: enforcement. The DOJ's prosecution logic against non-custodial developers, FinCEN's regulatory ambitions in the self-custody space, the SEC's stance on unregistered digital assets—none of these will pause because a bill dies. This is the most underestimated cost of "failure": not a neutral outcome, but active enforcement without the protection of statutory shields.

How Is the Industry Debating This Now?

It's not monolithic; there are at least six distinct voices.

The first, and currently most influential faction: Pessimistic, but believes it's a negotiation tactic.

Paradigm's Head of Government Affairs, Alexander Grieve, is widely quoted. He argues those calling the probability "close to zero" are misreading anonymous staffer briefings—when lawmakers express pessimism to reporters, it's to extract last-minute concessions from the White House or Senate leadership. His evidence: bank lobbying and ad money are still flowing; no one spends money on a corpse.

He acknowledges the schedule is indeed a problem, but adds:

"But this thing isn't dead, not at all."

The second, Coinbase's Brian Armstrong: Clarity will definitely come; legislation is just one path.

He said on CNBC that if the bill passes, the industry gets durable legislation; if not, the SEC and CFTC will still issue rules. His judgment is that "clarity could come on September 15 or in the days immediately after." He also said Coinbase's "must-have provisions" have been resolved, leaving the core disagreement as the ethics rules.

Background: In January, Armstrong temporarily withdrew his support, stating, "I'd rather have no bill than a bad bill."

The third, a group opposite Armstrong: Keep negotiating, don't walk away from the table.

Kraken CEO Arjun Sethi advocates resolving remaining issues through negotiation, not abandoning years of bipartisan progress. a16z Managing Partner Chris Dixon, Ripple CEO Brad Garlinghouse, and White House crypto advisor David Sacks hold similar positions.

Ledger's Global Head of Policy, Seth Hertlein, captures many people's sentiment: It's hard to imagine a more favorable environment in the future—if it can't be done now, it will either not get done later, or have to be done under very unfavorable conditions.

The fourth, academic and regulatory skeptics: The bill itself is problematic and should be revised, regardless of passage.

Columbia Law School's CLS Blue Sky Blog published "Shadow SEC Statement No. 14: Reject the Clarity Act" on September 10, giving four reasons for opposition: the ethics provision is so lax it seems to endorse past Trump family dealings; it won't bring clarity, instead favoring incumbents; the severely understaffed CFTC can't handle this job; the long list of SEC exemptions brought by network tokens and affiliated assets.

The fifth, law enforcement and consumer protection camp: Softening, but not over.

The National Sheriffs' Association (NSA) has shifted from "oppose" to "neutral." It previously called the bill "harmful" and BRCA language "bad policy." Now, no major police organization is publicly opposing it.

But prosecutor groups (NAAUSA, NDAA) still oppose BRCA, and Cortez Masto, a key vote on BRCA, has yet to publicly change her stance.

The sixth, direct clash between banks and crypto: Stablecoin yields.

The American Bankers Association continues to call for closing the "stablecoin interest and yield" loophole, with community banks worried about deposit flight. JPMorgan's Jamie Dimon opposes stablecoin yields, while Goldman Sachs' David Solomon publicly supports them. This front, alongside the ethics provision, is a reason Republicans might also defect.

Traders, on the other hand, have a clear attitude.

Over $3.6 million bet on "No," with two large positions built just before the vote; those long holding Yes are sitting on floating losses.

Industry discussions on X are another mix: "Clarity will come, whether it passes or not," interspersed with numerous bets on price movements after passage or failure.

Outside pressure is still ramping up.

Treasury Secretary Scott Bessent called on senators on X on September 9 to support the motion to proceed. His argument shifted angle—from "market structure" to national security, saying abandoning legislation would signal US unwillingness to lead in digital assets and could lose enhanced tools to combat abuse.

White House crypto advisor Patrick Witt urged all senators, regardless of party, to vote in favor first to allow the process to move forward.

On the industry side, political funds aimed at influencing congressional elections have surpassed $190 million. Ripple's Chief Legal Officer, Stuart Alderoty, has been directly calling offices of senators who haven't taken a stance or have voiced opposition, asking lawmakers to meet directly with ordinary token holders, not industry executives or lobbyists, citing "67 million Americans holding crypto assets."

Some Unanswered Questions

This revision was a technical fine-tuning. The four issues truly blocking the bill remain untouched: the ethics provision targeting Trump family crypto interests, the BRCA's developer liability exemption, stablecoin yields, plus the newly emerged this week: regulatory authority over vertical integration. The first three are old problems from July; the fourth is new.

So September 15 is more like a signaling vote: It doesn't decide the bill's fate; it decides who is on the list.

If passage means writing rules into law, and failure means handing rules to agencies, then the true stakes in this battle are whether, in the future, there will be a statutory shield in front of self-custody and developer liability.

Three questions cannot be answered now.

Will the 7 to 10 missing Democratic votes needed for that 60 materialize in the final 5 days?

The White House has been silent for over a month on the Tillis-Gallego bipartisan ethics proposal; will it speak up before the vote?

And—if this bill truly halts on September 15, will the industry accept the fragmented path of split bills, or will it, as in January, adopt a hardline stance of "we'd rather have no bill" once more?

Time will tell.

Related Questions

QWhat is the main purpose of the CLARITY Act (H.R. 3633)?

AThe primary purpose of the CLARITY Act, also known as the Digital Asset Market Clarity Act, is to clearly define and assign regulatory jurisdiction over digital assets between the SEC (Securities and Exchange Commission) and the CFTC (Commodity Futures Trading Commission), settling the decade-long debate on whether tokens are securities or commodities.

QAccording to the article, what is the key procedural vote scheduled for September 15th, and what is its significance?

AThe vote scheduled for September 15th at 2:15 PM EST is a cloture vote in the US Senate. This vote does not decide if the bill becomes law, but rather determines whether senators can begin formal debate on the CLARITY Act. Its significance is that if it fails to reach the required 60 'yes' votes, it is likely the practical end for the bill in 2026, due to the limited legislative calendar before the midterm elections.

QWhat are the four major changes the CLARITY Act would bring if passed, as outlined in the article?

AIf passed, the CLARITY Act would bring four major changes: 1) Clearly codify the jurisdictional split between the SEC and CFTC for digital assets. 2) Require intermediaries like digital commodity exchanges and brokers to register with the CFTC, imposing rules for asset segregation, custody, and market monitoring. 3) Provide legal safe harbors for non-custodial software developers (the BRCA) and self-custody wallet protections. 4) Implement an ethics clause restricting high-ranking officials and their spouses from issuing or sponsoring digital assets until 2029.

QWhat are the major sticking points, or controversial provisions, currently blocking the bill's progress according to the text?

AThe major sticking points blocking the bill's progress are: 1) The ethics clause, seen by some as too lenient and potentially tailored for past activities of figures like the Trump family. 2) The BRCA provision exempting non-custodial software developers from being treated as money transmitters. 3) The stablecoin yield provision, opposed by banks fearing deposit flight. 4) A newly emerged issue regarding vertical integration regulatory authority.

QWhat is the market's implied probability (as of the article's writing) of the CLARITY Act being signed into law in 2026, and based on what prediction market?

AAs of the article's writing, the prediction market Polymarket implies a probability of only 17-17.5% that the CLARITY Act (H.R.3633) will be signed into law in 2026. This probability dropped significantly from 82% in February and 65% before the August recess, following the delay of the key Senate vote.

Related Reads

OpenAI and Meta Join Forces, Bridging the "Last Mile" of Agent Commerce

On November 13, 2026, two major announcements landed the same day, together addressing the "last mile" of agent commerce: how AI assistants get through the door, and how they pay. Sierra and Meta released Poppy v1.0, a Personal Agent Protocol built on OAuth, HTTPS, OpenAPI, and the Model Context Protocol. Poppy lets businesses cryptographically verify three things: which assistant is making a request, which user it represents, and what permissions it holds. Companies publish a discovery file at /.well-known/poppy describing endpoints and allowed actions, while both user and business must consent to each interaction via two-way authorization. Shopify automatically enabled Poppy discovery files for merchants in 175+ countries, giving the protocol immediate scale, with over 100 design partners including Delta, 1Password, Okta, and Cloudflare. Separately, OpenAI launched ChatGPT Wallet for its personal assistant dots. The assistant can search, compare, and fill a cart, but must stop at checkout—users confirm via biometric authentication. Neither the assistant nor the merchant ever sees the real card number; card networks issue a one-time payment credential instead. Built with Amex, Mastercard, and Visa, the wallet runs entirely on traditional fiat rails—explicitly not crypto or stablecoins. It initially opens only to US Pro subscribers aged 18+. Notably, Visa and Mastercard span both initiatives, and OpenAI itself appears among Poppy's design partners—despite dots and Meta's Muse competing for the consumer entry point. The shared logic: consumer-facing competition, but shared infrastructure on the business side. The common bottom line is "human in the loop"—agent autonomy stops at the shopping cart. Open questions remain: liability when an assistant errs or is manipulated by malicious prompts, coverage beyond Shopify and US Pro users, and whether mandatory biometric confirmation for every small purchase will become friction that undermines the convenience promise.

marsbit48m ago

OpenAI and Meta Join Forces, Bridging the "Last Mile" of Agent Commerce

marsbit48m ago

Has the Time Come to Buy the Dip on Nvidia?

Nvidia announced on November 12 that Jensen Huang resigned as President, CEO, and Director, effective immediately. Shares fell as much as 14% intraday to $148.8, down nearly 40% from October highs, with over $600 billion in market value briefly erased. The founder's departure triggered the selloff, but the market is repricing the entire company rather than reacting to a single risk. Nvidia now trades at roughly 12x forward PE, its lowest in 15 years and below the S&P 500's ~18x, despite strong fundamentals: FY2026 revenue of $215.9 billion (+65%), data center revenue of $62.3 billion in a single quarter, ~75% gross margins, and Q2 FY2027 revenue of $96.2 billion with $21.3 billion free cash flow. Most investment banks cut targets but maintained Buy/Overweight ratings (Barclays $210, Goldman $220, Citi $225, RBC $230), implying ~20% upside, while Truist downgraded to Hold at $174. The consensus: the business is intact, but its valuation multiple needed resetting. Supporting demand, Huang forecast at GTC 2026 that Blackwell and Rubin AI chip demand would reach at least $1 trillion by 2027, double the prior 2026 estimate. However, risks remain beyond CEO succession. CUDA's lock-in effect is eroding on the inference side—Anthropic disclosed that its Claude Mythos Preview autonomously ported core GPU kernels to AMD flagship GPUs in three days with minimal human intervention, prompting investor Marcelo P. Lima's widely cited claim that "Nvidia's CUDA moat no longer exists." Supply bottlenecks lie in HBM and packaging, not wafers: TSMC can cut over 400 million GPU dies annually, but Nvidia ships only ~10 million GPUs yearly, constrained by HBM capacity (~3.75EB globally, ~70% allocated to Nvidia, ~260GB per GPU) and tight packaging, with HBM4 adjustments possibly delaying Rubin. Huang also leaves behind highly personalized assets—strategic technology judgment, supply chain negotiation with TSMC and SK Hynix, and direct channels to customers and governments—that interim CEO Colette Kress, the long-time CFO, may not fully replace. In short, the business remains, but the "faith premium" is being systematically stripped away. At 12x forward PE, much pessimism is priced in, offering a favorable risk-reward window if AI capex holds and demand materializes. Yet cheapness alone isn't sufficient: CUDA's erosion is slow but irreversible, HBM and packaging limit revenue upside, and succession uncertainty means the founder premium may not be fully removed. As Bernstein noted, "if you see two ants in the kitchen, there are more than two," warning that chains from scandal to governance vacuum to government probes could trigger further selloffs. Until the dust settles, what a Jensen-less Nvidia is worth remains an open question.

marsbit1h ago

Has the Time Come to Buy the Dip on Nvidia?

marsbit1h ago

AI Bull Market Trap? The Big Short Seller Strikes Again

Recent signals from Goldman Sachs, the Federal Reserve, and prominent short-seller Michael Burry are converging on concerns about the AI trade. Goldman warns that AI capital expenditure growth will likely peak in early 2027, with the boost to S&P 500 earnings falling from a 12-point peak in 2026 to a 5-point drag by 2028, shifting the market from a one-way rally to two-way volatility. The Fed has flagged AI as a potential threat to asset valuations and financial stability. Burry has doubled his Nasdaq 100 put options and rebuilt short positions in a semiconductor ETF, calling this a "bull market trap" analogous to the 2000 dot-com bubble, with Nvidia's influence a key worry. Four pressures weigh on the AI narrative: peaking capex growth, high interest rates that hit long-duration AI assets and leveraged GPU cloud providers like CoreWeave and Nebius, circular financing risks among interconnected tech firms, and tighter regulation under Democrats affecting data center energy, land, and permitting. Nvidia-related turmoil has shaken the pricing anchor for the entire supply chain. However, a valuation bubble bursting does not mean the AI industry is declining. Goldman notes US cloud giants trade at just 19.7x forward P/E, below their 20-year historical range and even below defensive staples, suggesting much pessimism is already priced in. Opportunities are shifting from upstream hardware "shovel sellers" toward fairly valued cloud giants with improving free cash flow, AI application companies delivering real ROI, and deeply discounted cyclical names like Korean memory. Conversely, GPU cloud providers dependent on Nvidia orders and cash-burning startups without products are most vulnerable. Investors should diversify, hedge, and focus on assets that convert spending into stable returns.

marsbit1h ago

AI Bull Market Trap? The Big Short Seller Strikes Again

marsbit1h ago

Just Offered 3.3 Million Yuan Salaries to Poach Talent, OpenAI Suddenly Dissolves Robotics Team—Is the Industry About to Change!

OpenAI has disbanded its standalone robotics research team, confirmed shortly after booking booth 208 at the CoRL robotics conference in Austin went empty on November 13. The move surprised many, given that just months earlier OpenAI was aggressively hiring robotics talent—some roles paying up to $500,000 base salary—and had expanded its San Francisco data-collection lab to about 100 workers across three shifts, with a second lab planned in Richmond, California. Researchers from the robotics team are now folded into Chief Research Officer Mark Chen's core research division, while robotics applications and hardware staff move to President Greg Brockman's Applied division. OpenAI says robotics remains part of its long-term vision and the reshuffle is meant to accelerate progress. This is not OpenAI's first robotics retreat: it ended its first independent robotics effort in 2020 to focus on large language models, which produced ChatGPT, and only rebuilt a team in 2024. OpenAI explains the latest pivot technically—researchers initially believed video-generation models like Sora could teach robots physics, but general models such as GPT-6 Astra proved capable of directly controlling robots, so robotics work is being merged into the general foundation-model team. The broader industry context supports this shift. Unitree's Wang Xingxing said his biggest mistake over the past decade was not trusting AI enough, noting GPT-6 integration exceeded expectations and that humanoid intelligence's tipping point may arrive within one to two years. Anthropic's Claude Fable 5.2 scored 64.9% on the RoboDojo Sim benchmark versus 29.8% for GPT-6 Astra, and Nvidia's Astranomy study showed GPT-6 Astra controlling simulated robots with 74.8% success using an agentic approach, beating the prior specialized-model best of 51.0%. Yet challenges remain: general models are costly and slow—GPT-6 Astra averaged 18.56 seconds per inference and about 30.9 million tokens (roughly $36.40 in API fees) per successful task—problematic for real factories where conveyor belts don't pause. Google DeepMind, Nvidia, Unitree and AgiBot continue training specialized robotics models; Google's Apollo 2 achieved 96% success sorting parts in Mercedes factories using paired reasoning and action models, while Nvidia's GR00T 2 reached 95% success after brief fine-tuning. Analysts suggest OpenAI now treats robotics as a downstream application of general intelligence, leveraging core model advances while using its data-collection and hardware capabilities to validate models on real robots.

marsbit1h ago

Just Offered 3.3 Million Yuan Salaries to Poach Talent, OpenAI Suddenly Dissolves Robotics Team—Is the Industry About to Change!

marsbit1h ago

Trading

Spot
活动图片