AgiBot (Zhiyuan)

Phase A — Understand the business

Company Overview

What it is. AgiBot (brand for Shanghai Zhiyuan Robotics, founded February 2023) is a vertically-integrated, full-stack embodied-intelligence company — not a robot-body vendor. It owns four layers: (1) hardware — bipedal + wheeled humanoids and dexterous hands (Yuanzheng/"Expedition," Lingxi, Jingling lines); (2) the embodied-AI "brain" — the Genie Operator foundation models GO-1 → GO-2; (3) the data pipeline — the open-sourced AgiBot World manipulation dataset + an "AIDEA" data factory of teleoperated demonstration data; (4) a deployment/operations layer (LinkCraft, Lingqu OS). HQ Shanghai; first assembly plant Jan 2024, a second Pudong line pushing capacity past ~400 units/month.

How it makes money (today) — three early motions:

1. Robot unit sales — humanoid + wheeled platforms into industrial, logistics, retail, hospitality, education, and reception/"explanation" use cases.
2. State-owned-enterprise (SOE) order book — >2 billion yuan (~$280–295M) of 2025 orders from Chinese SOEs for power plants, data centers, and entertainment. The most important and most ambiguous line in the company: it is an order figure, not recognized revenue, and it tethers demand to state procurement.
3. Software / models / data / dev-kits — GO-series models, the AgiBot World dataset (released open Dec 2024), educational robots, and OEM/operator agreements.

Reported revenue & the "358" plan (the strategic spine). At its April 17, 2026 partner conference (APC2026), Chairman Deng Taihua disclosed 2025 revenue of 1.05B yuan (~$154M), up from 60M yuan in 2024 — a ~17x jump the company claims makes it the fastest Chinese robotics firm to cross 1B yuan. The forward plan ("358"): >10B yuan (~$1.47B) by 2027 (the "second curve") and >100B yuan by 2030 (the "third curve"), with the 100,000th general-purpose robot targeted by end-2027. Deng's framing is unusually candid: "AgiBot is not in a rush to pursue profitability, staying focused on the long term" — i.e. a CEO explicitly telling you he is burning.

Customers / partners named (public): Fulin Precision Engineering (Aug 2025, ~100 Yuanzheng robots across its factories), Longcheer (tablet lines), Minth Group (full-portfolio launch in Germany, Mar 2026), Singtel Enterprise (first operator agreement, Singapore), plus unnamed SOEs (power/data-center/entertainment). International footprint claimed across Europe, North America, Japan, South Korea, SE Asia, the Middle East.

Why it matters. China shipped ~12,800 humanoids in 2025 (~90% of the global ~16,000) and commands ~80% of the global humanoid market. Western programs (Tesla, Figure, Boston Dynamics) own the capital and the attention; the Chinese duopoly — AgiBot + Unitree — owns the units. And the politics: in April 2025 Xi Jinping personally inspected AgiBot's robots in Shanghai, with Peng publicly pledging to "break down foreign technical barriers" — a top-of-state endorsement no Western peer has.

Supply Chain

Map: upstream components → AgiBot integration → end customer.

• Upstream (inputs): harmonic reducers / actuators, frameless torque motors, force/torque sensors, lidar + depth cameras, dexterous-hand tendons, and the chokepoint — advanced AI compute for training the GO models and on-robot inference. China's actuator/sensor base is deep and cheap (the structural cost advantage behind the whole duopoly), but high-end GPUs/edge-AI silicon sit under US export controls. Compute is the one genuinely foreign-exposed link; AgiBot has not disclosed its training-compute source. Notably, AgiBot is investing upstream into the actuator chokepoint — it took a stake in Nexforce, a core-actuator R&D firm, extending vertical integration into its own most strategic component.
• Midstream (AgiBot): in-house design + assembly across two Shanghai plants; an in-house data factory (AIDEA / training centers) generating the real-robot demonstration data that feeds GO-1/GO-2 — vertical integration pushed into the data layer, which most Western peers buy or scrape.
• Downstream (buyers): Chinese SOEs (the anchor), domestic industrial/logistics/retail operators, and a thin but growing export channel (Minth-Germany, Singtel-Singapore).
• Strategic vertical move — Swancor. The Swancor acquisition (anti-corrosion resins / composite materials, esp. wind-blade) is not an obvious supply-chain fit — it reads as a listing vehicle first, materials play second (Lens 5/11). Any genuine industrial logic would be securing composite/structural materials for at-scale manufacturing; the more likely read is a backdoor to a Shanghai/STAR listing.

Chokepoints: (1) advanced compute (export-control single-point risk); (2) the SOE order channel (concentration — Lens 13); (3) teleoperation/data-labelling labor for the data factory (commoditizing, low risk). The Chinese actuator/sensor base is the opposite of a chokepoint — it is the moat (Lens 3), and AgiBot is buying deeper into it (Nexforce).

Competitive Advantages (moats)

Durable advantages — graded:

• China cost + supply base (STRONG, but shared). Domestic actuators, reducers, sensors at a fraction of Western BOM, plus state subsidies — buyer rebates, subsidized land/rent, and a $138B national VC guidance fund that names embodied AI a priority. This is the structural reason a Chinese firm ships 10,000 units while Figure delivered ~350. It is real but shared with the entire Chinese field, not proprietary.
• Full-stack integration incl. data (MODERATE — and the genuine differentiator). AgiBot's thesis, the line Peng repeatedly draws against Unitree: owning the body and the GO foundation model and the demonstration-data factory and the deployment OS compounds in a way a body-only vendor can't. GO-2's hard benchmark wins — LIBERO 98.5% (1st across spatial/object/goal/long), LIBERO-Plus 86.6% zero-shot, VLABench 47.4, sim-to-real 82.9% on simulation-only training — are the first evidence the software is competitive, beating π0.5 and NVIDIA GR00T on those tests. The AgiBot World dataset was an IROS 2025 Best Paper finalist. But VLA foundation models are the least durable layer — the frontier moves monthly (Physical Intelligence, Skild, NVIDIA GR00T N1.6/N1.7, Figure Helix).
• Production-ramp learning curve (MODERATE). 1,000 (~2 yrs) → 5,000 (~1 yr) → 10,000 (3 months) — the last leg a >4x acceleration. That manufacturing-execution flywheel compounds (yield, supplier terms, deployment playbooks).
• Brand / founder distribution (WEAK-BUT-REAL in China). Peng Zhihui ("Zhihui Jun") is a Bilibili-famous "Wild Iron Man" with real pull in Chinese tech-talent and consumer attention — a recruiting + PR moat.
• Political capital (REAL, and double-edged). The Xi inspection + 15th-Five-Year-Plan blessing is a domestic demand-and-funding moat — and the precise feature that turns AgiBot into a US-scrutiny target (Lens 10/13).

Bargaining power. Over suppliers: high (deep, fragmented Chinese base; AgiBot is a large, scaling buyer, now also an upstream investor). Over customers: low-to-negative on the SOE channel — when anchor demand is state procurement with ~10% buyer rebates, the government effectively sets the clearing price. Power over private/export customers is unproven.

Net: the moat that matters is the full-stack + proprietary-data thesis; the cost moat is a Chinese-field commodity, not an AgiBot edge.

Segments

our figures is empty — AgiBot is private and discloses no segment P&L. Reconstructed qualitatively (n/a — private, not disclosed for all hard numbers):

Geographic: overwhelmingly China; export is a 2026–27 narrative, not yet a number. The honest read: there is no segment economics to analyze yet — only a contested top-line (1.05B yuan), an unrecognized order book (>2B yuan), and a stated trajectory (10B by 2027). That opacity is the finding.

---

Phase B — Measure performance

Funding & Valuation Trajectory (+private swap for "Earnings Result")

Round history (all, unaudited; Chinese rounds frequently undisclosed on amount):

Total disclosed raised: ~$84.2M across 3 reported rounds, 17 investors (Tracxn) — but a ~15B-yuan post-Series-B valuation implies far more capital absorbed than $84.2M, so the public raise figure is incomplete/unreliable for a Chinese company.

Burn signals. Not disclosed — but now management-confirmed in direction: Deng's "not in a rush to pursue profitability… saturated R&D investment" is an explicit statement of heavy burn. The read-across is brutal: Unitree — the closest comp, and one that is profitable — saw Q1-2026 adjusted net profit fall ~52% (to 40.3M yuan) purely on R&D + sales expansion even as revenue rose 68%. AgiBot, which has never disclosed a profit and is spending harder across a wider stack, is near-certainly loss-making at the net line. The 1.05B-yuan revenue is a scale flex, not a profit signal.

Valuation conflict (must surface): the latest private mark is ~$2.1B (Mar 2026); the HK IPO target is HK$40–50B (~$5.14–6.4B). That is a ~2.5–3x step-up the IPO must clear — the single biggest number-vs-number tension in this dossier. The IPO target asks public markets to pay roughly a Unitree-equivalent valuation for a company that, unlike Unitree, has shown no profit and no audited margin.

Founder / Narrative Sentiment Trend (+private swap for "Earnings Calls")

No earnings calls. Proxy = founder/management public posture across 2024 → 2026:

• Dec 2024: "mass production begun," 962 units as of Dec 15 — proof-of-capability register.
• Apr 2025 (Xi visit): Peng — Xi showed "great attention and care"; AgiBot would "break down foreign technical barriers" and "enhance the global competitiveness of China's embodied intelligence" — national-champion register.
• Early 2025 (SCMP): mgmt — "deploy new products in industrial scenarios, replacing humans… to create tangible customer value" — ROI/commercial register.
• Mar 2026 (10,000th unit): Peng — "Reaching 10,000 units… reflects a fundamental shift in our ability to scale." — scale/industrialization register.
• Apr 2026 (APC2026): Deng/Peng disclose the 1.05B-yuan flex, the 358 plan (10B 2027 / 100B 2030), and "not in a rush to pursue profitability"; Peng draws the explicit line — "Unitree focuses on the robot body, while ZhiYuan has a full-stack layout"; 4 new bodies + 6 AI models released.

Tone shift: spectacle (2024) → national-champion + commercial-ROI (2025) → scale + full-stack-differentiation + revenue-leadership + explicit profit-deferral (2026). The phrases they added: "full-stack," "tangible customer value," "fastest to 1B yuan," "second/third curve." The thing they now say out loud — that profit is deferred. The thing they still never quantify: gross margin or the size of the net loss. That selective silence is the tell.

Cap Table & Secondary Marks (+private swap for "Comps")

Syndicate quality — strong, and an IPO-proximity tell. Reported cap table: founders ~48%, ESOP ~10%, Tencent ~7%, then a long tail (4.5% → 1.5% each). Investor roster spans strategic corporates (Tencent, BYD, LG Electronics, JD.com/JD Technology, Baidu, SAIC Capital, Wolong, Longcheer, TCL), top-tier VC/PE (HongShan/ex-Sequoia China, Hillhouse, CDH, Warburg Pincus, BlueRun, Lanchi), and crossover/foreign capital (LG, Mirae Asset, Warburg Pincus). A Warburg/Mirae-grade crossover presence is exactly the late-stage signature that precedes a public listing.

Peer table (the cleanest comp is Unitree — it has an audited STAR-Market prospectus, cleared 2026-06-01):

Shipment-data conflict (flag): Omdia (Jan 2026) ranks AgiBot #1 at 5,168 / Unitree 4,200 for 2025; "who's #1" is definitionally contested (humanoid-only vs all embodied units; ship vs produce). Do not treat AgiBot's #1 claim as settled.

Read: against the only audited peer, AgiBot has less revenue, no proven profit, no disclosed margin — yet seeks a richer EV/Sales multiple. Unitree is the benchmark the IPO will be priced against, and Unitree clears it on profitability + 60% gross margin where AgiBot today cannot. (Caveat the other way: Unitree's own ~150x P/E shows the public market is already paying dream multiples in this group — which is exactly what makes AgiBot's listing feasible even if demanding.)

Funding / Product Catalysts (+private swap for "Stock-Price Catalysts")

Events that re-rated AgiBot's private mark / narrative:

• Mar 2025 — GO-1 launch (first generalist embodied foundation model, ViLLA architecture) — established the software story.
• Mar 24, 2025 — Tencent-led Series B → unicorn at ~15B-yuan post.
• Apr 2025 — Xi Jinping inspection in Shanghai — the national-champion catalyst (political, not financial).
• Jul 2025 — Swancor stake bid (~2.1B yuan / ~$290M, 63.62%) — the listing-vehicle catalyst; makes AgiBot's economics partially tradeable via Shanghai-listed Swancor.
• Aug 2025 — Fulin Precision deal (~100 robots) — a named commercial deployment.
• Nov 10, 2025 — LLC → joint-stock conversion — the standard pre-IPO restructuring step.
• Mar 30–31, 2026 — 10,000th unit (Expedition A3) — the scale-leadership catalyst.
• Apr 9, 2026 — GO-2 launch (beats π0.5 / GR00T on benchmarks) — software-credibility catalyst.
• Apr 17, 2026 — APC2026: 1.05B-yuan revenue + the 358 plan + 4 bodies/6 models.
• 2026 (rumored) — HK IPO, prelim prospectus early 2026, listing ~Q3 2026, target ~$5.1–6.4B, banks CICC, CITIC Securities, Morgan Stanley.

Pattern: AgiBot's marks move on (a) tier-1 capital entries, (b) production milestones, (c) the listing machinery, and (d) state endorsement — not on disclosed financial performance (there is none to react to). The market is pricing scale + story + a near-term liquidity event — precisely the risk profile of a momentum pre-IPO name.

---

Phase C — Judge people & books

Management

• Deng Taihua (邓泰华) — Chairman & CEO; de-facto controller. A former Huawei VP (computing / Ascend-adjacent business lineage). The institutional operator and named de-facto controller of the Swancor vehicle, pledged to hold control of the listed entity. He, not Peng, is the corporate/financial principal — and the author of the public "358" plan and the explicit profit-deferral.
• Peng Zhihui (彭志辉 / "Zhihui Jun") — Founder, CTO/President. Born 1993; joined Huawei in 2020 via the "Top Minds"/"Genius Youth" programme for AI; UESTC master's; Bilibili-famous DIY-robotics creator; prior 3D-printing startup. The technical visionary, public face, and recruiting magnet. Young (~32) — first-time scale-operator; the CEO/CTO split with the older Huawei VP is the governance design that mitigates this.
• Other founders: technical co-founders include academics from Shanghai Jiao Tong University.

Track record: genuine — built a 10,000-unit humanoid business and a benchmark-competitive foundation model from a Feb-2023 standing start in three years. Capital-allocation history (thin, three notable moves): (1) extending vertical integration into a data factory (defensible, on-strategy); (2) upstream into actuators (Nexforce stake) — on-strategy, secures the chokepoint; (3) the Swancor acquisition — capital into a wind-blade/composite-materials firm whose strategic fit is, charitably, indirect. If Swancor is a listing vehicle dressed as M&A, it is a governance/transparency yellow flag, even if legal.

Founder vs professional manager: a deliberate hybrid — founder-technologist (Peng) + professional-executive controller (Deng, ex-Huawei VP). For a capital-intensive hardware company chasing a public listing, the professional-CEO anchor is the right archetype; the open question is whether the Huawei-VP lineage is a competence asset or a Huawei-/state-adjacency liability in Western markets (export, procurement scrutiny). Both are true at once.

Forensic Red Flags + Regulatory

Accounting / disclosure risks (private → mostly un-testable, which is itself the risk):

• Revenue-recognition opacity (HIGH concern). "1.05B yuan, +17x" cannot be reconciled with a third-party ~$10.8M read. With no audited statements, we cannot tell how much of "revenue" is hardware sell-through vs the >2B-yuan SOE order book vs government/data-factory contracts. A 17x jump coinciding with a state-procurement surge and a pending IPO is exactly the pattern a forensic analyst flags for order-vs-revenue conflation / channel-stuffing / related-state-party — not as an accusation, but as the un-verifiable zone that audited IPO statements must clear.
• SOE-order concentration & relatedness. >2B yuan from state buyers, in a sector receiving heavy subsidies + ~10% buyer rebates, blurs commercial revenue and state-directed demand. Quality-of-revenue is the central diligence question.
• Swancor backdoor / control mechanics. AgiBot denies the deal meets China's legal definition of a backdoor listing (per the Administrative Measures for Major Asset Restructurings); that it had to deny it is the flag. The structure — 29.99% via negotiated transfer + 37% via partial tender = 63.62%, controlling shareholder shifting to a Deng-controlled partnership, after an LLC→joint-stock conversion — is the textbook pre-listing reorganization. Legal ≠ transparent.
• R&D capitalization / inventory build. A 10,000-unit ramp with thin disclosed sell-through raises finished-goods-inventory and capitalized-development questions — unanswerable without statements, but the place a short would dig (units produced ≠ units sold/recognized).
• Management self-disclosure cuts both ways. Deng openly saying profit is deferred is more honest than most pre-IPO promotion — but it also confirms the prospectus will show a loss, so the IPO must be sold on the trajectory, not the P&L.

Regulatory findings. Per regulatory/regulatory-findings.md (fetched 2026-06-30): AgiBot has no CIK; no SEC EDGAR enforcement (LR/AAER) is possible — 0 SEC findings. Non-SEC web search ("AgiBot"/"Zhiyuan Robotics" + FTC/DOJ/FDA/CFPB/consent-decree/settlement/fine/penalty) returns no material enforcement actions, lawsuits, or penalties as of this date. No 10-KForm 10-KA company’s audited annual report to the US regulator. The most complete thing it publishes. Item 3 exists (private/foreign). The live regulatory exposure is forward-looking, not historical, and it has hardened in 2026:

1. China CSRC/SSE review of the Swancor restructuring and any HK listing.
2. US export-control / federal-procurement risk — now concrete, not speculative. The US Senate is considering a (limited) federal procurement ban on Chinese robotics; CSIS and Hudson are explicitly urging targeted export controls on sensors/software used by Chinese humanoids; and China's National Humanoid Robot Standardization Committee embeds PLA priorities and US-restricted firms at its core. Given AgiBot's Huawei-VP lineage, SOE customers, Xi endorsement, and dual-use humanoid tech, an entity-list designation or procurement ban is the single most plausible future enforcement vector — currently unrealized but trending the wrong way.

Net: no material historical regulatory or legal findings — verified via SEC EDGAR EFTS (none possible), web search, and the absence of any required public filing, as of 2026-06-30; all findings unaudited per public sources. The risk is entirely forward-looking and geopolitical.

---

Phase D — Project & stress-test

IPO-Readiness & Path-to-Tradeable (+private swap for "Forward Projection")

Readiness: 4–5 on the private-watch scale (pre-IPO / listing imminent). The milestones that unlock a listing are largely done or in motion:

• ✅ LLC → joint-stock conversion (Nov 10, 2025) — the structural prerequisite.
• ✅ Tier-1 + crossover syndicate (Tencent, Warburg, Mirae, LG) — book-building credibility.
• ✅ Swancor 63.62% control → a Shanghai-listed (STAR-eligible) vehicle already partly tradeable.
• ✅ Banks engaged — CICC, CITIC Securities, Morgan Stanley.
• 🟡 HK IPO: prelim prospectus targeted early 2026, listing ~Q3 2026, 15–25% of shares for >$1B at HK$40–50B (~$5.1–6.4B).
• 🟡 Audited financials: the gating item — the IPO is the first time AgiBot's revenue composition, margin, and (likely) loss face an auditor and a prospectus.

Two parallel paths to "tradeable": (1) Swancor / Shanghai-STAR — already live, partial exposure now; (2) HK IPO — the clean, full-float event, ~H2-2026. Base-case window for a clean tradeable AgiBot: Q3–Q4 2026, with execution + China-market-window + audit-disclosure the three things that slip it.

Valuation framing (no EPS — pre-profit; only). At the IPO target ~$5.1–6.4B on ~$154M (contested) 2025 revenue → ~33–42x EV/Sales — vs Unitree's audited ~25x with a real 60% gross margin and 278M-yuan profit. The forward story is the 358 plan: >10B yuan by 2027 (a ~10x in two years off 2025). If even half-delivered (~5B yuan / ~$700M), the IPO multiple compresses to ~9x forward sales — if you believe the trajectory and if margins ever appear. The honest base case: the listing happens in 2026; the price is the bet, and at the rumored level it is demanding relative to the only audited peer. AgiBot must (a) prove revenue is higher/cleaner than skeptics think, OR (b) grow into the multiple at the 358-plan pace, OR (c) price below the rumored target.

No Brier forecast logged (--watchlist rule: skip our model create). If promoted to a thesis, the scoreable binary is "AgiBot completes a public listing (HK IPO or Swancor-STAR control transaction closes) by 2026-12-31," base-case p ≈ 0.70. A second, harder binary worth tracking: "AgiBot's IPO prospectus discloses a 2025 net loss," p ≈ 0.80.

Recommended overlay write-back (NOT executed under wave boundaries): add AgiBot to research/private-watch.json at readiness 4–5, beat robotics, lead investors "Tencent, HongShan, Hillhouse, Warburg," catalyst "HK IPO H2-2026 (CICC/CITIC/MS) + Swancor-STAR control deal; 358 plan to 10B yuan 2027," dossier:../menfem-research/companies/agibot/the previous dossier. Flagged for Connor; not written (wave boundary). AgiBot is currently absent from private-watch.json — this would be the entry creation.

Bull vs Bear

Bull case. AgiBot is the purest, most-scaled pre-IPO embodied-AI bet investable, on the right side of three structural forces: (1) China's heavily-subsidized, 15th-Five-Year-Plan-blessed, Xi-endorsed embodied-intelligence build-out, where it is the top-2 volume player (Omdia #1, 2025); (2) a genuine full-stack + proprietary-data differentiation that GO-2's benchmark wins now substantiate (98.5% LIBERO, 82.9% sim-to-real beating π0.5/GR00T), in a field where most "humanoids" still don't run foundation models; (3) a credible, near-term liquidity event (HK 2026 + Swancor) backed by a Tencent/Warburg/Mirae-grade syndicate and top-bracket banks. If even the lower-bound humanoid TAM lands and the Chinese duopoly holds ~80% domestic share, the volume champion with the best brain compounds — and the 358 plan (10B 2027 / 100B 2030) is the option you're really buying. Contrarian view the market refuses to see: the West mis-prices this as "another cheap Chinese Unitree clone" when the data-factory + GO-model stack may be the more durable asset than the body — and the SOE order book, however murky, is a demand floor Western peers don't have.

Bear case (permanent-impairment risks).

1. The revenue is half a mirage. If audited IPO statements reveal the bulk of "1.05B yuan" is unrecognized SOE orders, one-time state/data-factory deals, and that the company is deeply loss-making (Deng has already conceded profit is deferred; Unitree's 52% profit collapse on a profitable base shows how brutal the spend is) — the equity story breaks at the prospectus.
2. Commodity economics + price war. 150 Chinese humanoid firms, state-subsidized overcapacity (an NDRC official has publicly warned of it), ~10% buyer rebates, Unitree shipping an R1 at $5,900 → a structural price war that caps gross margin below the level needed to ever earn the multiple. Volume leadership in a deflating-ASP commodity is not a profit pool.
3. Compute / export-control choke + geopolitics. Training frontier VLA models needs advanced compute under US export control; the Huawei-VP lineage + SOE ties + Xi endorsement make AgiBot a plausible entity-list / federal-procurement-ban target (now actively debated in the US Senate) — capping the very international growth the bull case needs.

Pre-mortem (18 months out, thesis broke): AgiBot IPO'd in late 2026; the prospectus showed a large net loss and revealed ~half of 2025 "revenue" was SOE order recognition; the stock priced 30–40% below the HK$40–50B target; a domestic price war crushed unit margins; and a US procurement ban + sensor export controls shut the international channel. Volume kept rising, profit never came, the multiple compressed toward Unitree's level on a worse margin profile.

Are multiples too high? At the rumored ~$5.1–6.4B / ~33–42x sales for a pre-profit company, yes — relative to the only audited peer. The bet is the re-rate the IPO would require and the 358 trajectory, not the current private mark.

Devil's Advocate (short-seller)

Dismantling the bull case:

• Where revenue is concentrated: the SOE/state channel (>2B yuan of 2025 orders). If state procurement slows, redirects to a rival, or proves to be subsidized demand-pull rather than commercial pull, the growth story is hostage to Beijing's budget and industrial-policy whims — not a customer base, a policy beneficiary.
• Why the moat is weaker than bulls think: the cost advantage is the whole Chinese field's, not AgiBot's; the VLA-model edge (GO-2) is the fastest-decaying layer in tech — π0.5, GR00T N1.6/N1.7, Figure Helix, Skild iterate monthly, and an April-2026 benchmark lead is worth little by 2027. Body + actuators are commoditizing under a 150-player price war. What, exactly, is durable in 3 years?
• Most dangerous competitor bulls underestimate: Unitree — already public, profitable (~278M yuan, 60% gross margin), lower cost structure, raising ~$616M to out-spend AgiBot on R&D while AgiBot bleeds. Plus UBTech (HKEX-listed, $1B financing facility, BYD deployment = world's largest commercial humanoid install) and a long subsidized tail.
• Worst capital-allocation / governance move: the Swancor backdoor — buying a wind-blade materials firm with no clear robot-supply logic, denying it's a backdoor listing, installing a de-facto controller. Legal or not, it signals the priority is the listing, not the operating logic — and minority public shareholders inherit that structure.
• Assumptions that must hold for the rumored price: (a) revenue is real, recurring, commercial (not order-book/state); (b) margins reach Unitree-like 60% despite a price war; (c) the international channel scales without an export-control/procurement block; (d) the VLA edge persists; (e) the 358 plan is roughly on track. If 2027 revenue disappoints 20–30% off the 10B-yuan target and margin lands at ~30% in a price war, a ~33–42x sales multiple compresses toward 15–20x — a ~50% valuation cut from the IPO target.
• Single permanent-impairment scenario: a US entity-list designation + procurement ban (Huawei-lineage + SOE + Xi-endorsed + dual-use) that severs advanced-compute access and the export channel simultaneously — plausibility moderate and rising given the trajectory (active Senate debate; CSIS/Hudson advocacy).

Management Questions (ordered by information value)

1. Of the ~1.05B-yuan 2025 revenue, what is the GAAP-recognized split between hardware sell-through, the SOE order book, government/data-factory contracts, and software — and what was the audited net loss? (The whole thesis hinges on this single answer.)
2. What was 2025 gross margin, and where does it go in 2026–27 given the 150-player price war and ~10% buyer rebates?
3. What fraction of 2025 demand was state-directed / subsidized vs. independent commercial, and how do you de-risk dependence on SOE procurement?
4. What is your monthly cash burn and runway, and does the planned IPO raise (>$1B) cover the path to the 358-plan inflection — or just to the next raise?
5. Justify the HK$40–50B target against Unitree's ~$6.2B with audited ~$250M revenue, 60% margin, and real profit.
6. The 358 plan implies ~10x revenue in two years (to 10B yuan by 2027). What is the bottom-up bridge — units × ASP × mix — and how much rests on SOE/government vs. commercial vs. export?
7. What is your source of advanced training compute, and your contingency if export controls tighten on the silicon you depend on?
8. How do you assess US entity-list / federal-procurement-ban risk given the founding team's Huawei lineage, SOE customers, and the Xi endorsement — and what is the mitigation for the international growth plan?
9. What is the strategic logic of the Swancor acquisition beyond a listing vehicle, and what do public minority shareholders get from the de-facto-controller structure?
10. Of units produced (10,000+, targeting 100,000 by end-2027), how many are deployed and generating recurring revenue vs. inventory, pilots, or unpaid demos?
11. What is the attach rate and pricing of GO-series models / software per robot — is software a revenue line or a giveaway to sell hardware?
12. How durable is the GO-2 benchmark lead when π0.5, GR00T, and Helix iterate monthly — what is the proprietary, non-replicable component (data? the AIDEA factory?)?
13. What is the economic moat in 3 years if actuators, sensors, and VLA models all commoditize across 150+ Chinese firms?
14. What governance and capital-allocation guardrails change as a public company (related-party SOE sales, controller lock-up, M&A discipline)?
15. How do the CEO/CTO decision rights divide between Deng (controller) and Peng (founder-technologist), and what is the founder retention / lock-up post-IPO?

---