· Updated · Venture Capital Tracker · investment-strategies · 3 min read
Sharpa’s ~$670M Raise: Hesai Founders Bet on Dexterous Hands, Not Lab Demos
Shanghai robotics startup Sharpa disclosed more than ¥4.5B (~$670M) with Alibaba, Tencent, Meituan, JD, and HongShan, at a reported ¥22B post. Mass-produced Wave hands and a live Dairy Queen robot restaurant are the proof claims — not a single /fund/ lead on this site.
VCT data record
Funding event facts
Source-backed financing and transaction details. Unknown terms remain undisclosed rather than estimated.
Sharpa raises more than ¥4.5B (~$670M)
First public financing disclosure. Alibaba, Tencent, Meituan, JD among strategics. Reported ~¥22B post.
- Event type
- Funding Round
- Event date
- Aug 28, 2026
- Stage / label
- Growth
- Amount
- ~$670M
- Valuation
- ~¥22B post (reported)
- Confidence
- Reported
Company / target: Sharpa
Participants: Alibaba Group , Tencent , Meituan , JD.com , HongShan
Sources: yangtzeer.com dealstreetasia.com
Sharpa disclosed more than ¥4.5 billion (~$670 million) on August 28, 2026 — its first public financing print — at a reported ¥22 billion post. Alibaba, Meituan, Tencent, JD.com, and Transsion sit next to HongShan, Qiming, Meituan DragonBall, and Luminous. None of those names have a /fund/ page here.
Unexpected truth: the story is not “another humanoid raise.” It is a dexterous-hand company from the Hesai founders, shipping a 22-DoF Wave hand into mass production and putting robots on a live Dairy Queen shift — then raising at consumer-internet scale.
This page is for founders and scouts mapping China’s physical-AI capital. Last verified August 31, 2026. English coverage is secondary (Yangtzeer, DealStreetAsia); company WeChat is the primary Chinese disclosure path.
Five-minute decision
| If you need… | Verdict |
|---|---|
| What happened | Funded (reported). >¥4.5B / ~$670M. ~¥22B post (secondary). |
| What Sharpa is | Dexterous hand (Wave) + wheeled humanoid (North) + CraftNet manipulation model. |
| Whether it is shipping | Company: Wave mass production since Oct 2025; DQ robot restaurant live Aug 29 path. |
| Whether to diligence | Yes, if you underwrite China’s manipulation bottleneck. No, if you need a US /fund/ lead. |
Investigate further when: you need a named component supplier behind GR00T-class demos.
Wait or pass when: you need audited revenue or a priced round filing in English.
What happened
| Field | Detail |
|---|---|
| Company | Sharpa (Shanghai; founded late 2024) |
| Founders | David Li Yifan, Xiang Shaoqing, Sun Kai — Hesai Technology co-founders |
| Amount | >¥4.5B (~$670M) · first public disclosure · Aug 28, 2026 |
| Valuation | ~¥22B post (reported) |
| Strategics | Alibaba, Meituan, Tencent, JD.com, Transsion |
| VCs named | HongShan, Qiming, Meituan DragonBall, Luminous |
| Products | Wave (22-DoF hand), North (wheeled humanoid), CraftNet (manipulation model) |
Who uses it — and why they would
Buyers / users (company and press):
- Robot OEMs and reference platforms that need a high-DoF hand (Nvidia Isaac GR00T reference path: Unitree body + Sharpa hands, per Yangtzeer).
- Franchise / retail operators testing unmodified store automation (Dairy Queen robot restaurant: multi-step Blizzard prep without store retrofit — company/press).
- Industrial teams that treat manipulation — not locomotion — as the scarce skill.
Why Alibaba / Tencent / Meituan / JD write this check (judgment, labeled): platform cos already own distribution, cloud, and last-mile demand. Dexterous robots that finish the “last 1%” of physical work turn those surfaces into automation buyers — and generate proprietary contact data. HongShan/Qiming supply growth-VC process around that strategic stack. This is a portfolio-fit story for China tech conglomerates, not a classic US seed syndicate.
Why now
China’s dexterous-hand market is in a capital sprint: industry estimates cited by Yangtzeer put domestic-sector funding above ¥8B by end-July 2026 vs ~¥500M a year earlier. Average hand prices falling from ~¥50k to below ¥10k changes the BOM math for fleet deployments. Sharpa’s print arrives the day before a public DQ restaurant demo — marketing and financing in the same 48 hours.
What is not proven
- English sources do not break primary vs secondary capital.
- No ARR, unit shipments, or gross margin in the coverage used here.
- Restaurant demos ≠ multi-city uptime SLAs.
- LinkerBot and others still claim large shares of the high-DoF hand market — Sharpa has not published a bake-off.
Competitive map
| Player | Angle | Sharpa contrast |
|---|---|---|
| LinkerBot / Inspire / BrainCo | Independent dexterous-hand specialists | Sharpa bundles hand + whole-body + model + retail proof |
| Unitree / AgiBot | Robot OEMs building own hands | Sharpa also supplies hands into others’ platforms |
| US humanoid stacks (Apptronik et al.) | Western capital + different go-to-market | Sharpa’s check writers are China platforms |
Implication
The unresolved question is whether a ¥22B mark prices a components company or a full-stack Physical AI platform — and whether DQ-style shifts become paid, multi-site contracts before the next mark. Until that lands, treat the round as a strategic capacity bet on manipulation, not a SaaS multiple.
Follow Venture Capital Tracker in Google
Add VCT as a preferred source to make our venture-capital coverage easier to find in Google Search.
Last updated:
Editorial note: AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.
