The Valuation Debate: When a Profitable Company Gets Called “The Next Tesla”
Unitree valuation has reportedly reached $70 billion — making it one of the most valuable robotics companies on earth. But does the business justify the number?
In mid-2025, as Agibot moved to go public via a reverse merger — acquiring a listed Chinese industrial materials firm for RMB 2.1 billion (USD$300M) — the market began pricing humanoid robotics as a winner-takes-all race.
Agibot’s stock surged 330% in eight days.
Other robotics firms rushed to file IPO paperwork.
And suddenly, Unitree Robotics — a company that had never disclosed a valuation — was being whispered about as a RMB 50 billion (USD$7 billion) candidate.
Then came the denial.
Unitree’s leadership publicly stated: “We have not discussed any IPO valuation. Any figures circulating are untrue.”
That wasn’t a rejection of the market’s expectations.
It was a signal: They’re not here to be priced by hype.
Unitree Valuation: The Numbers That Matter
Unitree is not a startup burning cash.
It is a profitable, export-driven hardware company with a clear, repeatable business model.
| Metric | Value |
|---|---|
| 2024 Revenue | RMB 1.05 billion (USD$145 million) |
| Profitability | Five consecutive years of net profit |
| Gross Margin (Robot Division) | 45–55% |
| Primary Product | Quadruped robots (Go2, G1 series) |
| Quadruped Market Share (Global) | ~70% |
| Export Revenue | 50% of total sales — primarily to U.S., Europe, Japan |
| Humanoid Revenue | ~30% of total — all non-industrial (education, research, consumer) |
Key insight:
Unitree’s profitability does not come from humanoid robots.
It comes from selling thousands of low-cost, high-margin quadruped robots — priced as low as $1,400 USD (RMB 9,997), versus Boston Dynamics’ $538,000 Spot.
This is not a “moonshot” company.
It’s a high-volume, high-margin precision hardware play — the DJI of robot legs.
The Humanoid Problem: Performance ≠ Profitability
Unitree’s humanoid robot — the G1 — is technically impressive:
- Lightweight, agile, stable
- Lower cost than most competitors
- Built on proven hardware architecture
But here’s what it cannot do:
- Operate autonomously in unstructured environments
- Learn new tasks without manual programming
- Solve industrial problems better than a fixed-arm robotic arm or a wheeled AGV
Its use cases?
- University labs
- Science museums
- Tech demos
- YouTube videos
No factory. No warehouse. No hospital. No commercial contract.
In contrast:
- Agibot has deployed robots in automotive assembly lines for part handling.
- Tesla Optimus is testing in its own Gigafactories.
- Figure AI is collecting real-world data from BMW’s production lines.
Unitree’s humanoid remains a research tool — not a productivity tool.
And that matters.
📉 The Real Risk: Being the Best Hardware Maker in a Software-Driven Race
The industry is shifting.
The next 10 years won’t be won by the company with the best joints.
They’ll be won by the company with the best brain.
| Agibot | Task-specific AI + factory integration | $400–500M/year in AI models |
| Tesla | FSD chip + end-to-end learning | $1B+ in AI/robotics R&D |
| Figure AI | Data闭环 (real-world + home) | $300M+ in simulation + behavioral AI |
| Unitree | Hardware cost, reliability, scale | ~20% of R&D to software |
Unitree’s total R&D spend over the past three years: ~RMB 350 million ($48 million).
Of that, less than RMB 70 million ($10 million) went into AI models.
Agibot spends more per year than Unitree has spent in three.
The math is simple:
If you build the best robot body — but outsource the brain — you become a supplier, not a platform.
And suppliers get commoditized.
🚀 The IPO Isn’t About Money — It’s About Time
Unitree’s IPO filing isn’t about raising capital to expand production.
It’s about raising capital to buy time.
Why?
Because the window for hardware-first robotics is closing.
- Capital is flooding into the space: RMB 27 billion ($3.7B) raised in China alone in H1 2025.
- Valuations are decoupled from fundamentals: Some firms are valued at 10x revenue — despite no revenue from real applications.
- AI models are accelerating: The gap between “can walk” and “can work” is shrinking — fast.
Unitree’s goal isn’t to be the biggest robot maker.
It’s to become the first hardware leader to transition into a software-enabled platform.
That requires:
- $1–2 billion in AI R&D over the next 5 years
- Partnerships with cloud providers (AWS, Azure)
- Open access to its data and simulation tools
- Hiring AI researchers — not just mechanical engineers
The IPO is a bridge — from “hardware exporter” to “platform builder.”
Without it, they risk becoming the NVIDIA of the early 2020s — brilliant chips, but overshadowed by the companies that built the AI on top.
🔍 The Bigger Picture: Who Wins When the Hype Fades?
The robotics market is entering its “shakeout phase.”
- ~286 funding rounds in China in H1 2025 — up 138% YoY
- ~$3.7B raised — but only a handful have real revenue
- Only 3 companies have demonstrated repeatable, paid deployments in industrial settings
When the money dries up — and it will — the survivors won’t be the ones with the flashiest demos.
They’ll be the ones with:
- Proven ROI (e.g., reduced labor cost per task)
- Scalable software (not just firmware)
- Defensible IP (not just assembly lines)
Unitree has two advantages:
- Profitability — it doesn’t need to burn cash to survive
- Global distribution — it already sells to 50+ countries
But it has one critical vulnerability:
It hasn’t yet proven that its robots can do more than walk — they must learn.
Final Takeaway: Is the Unitree Valuation Justified?
Unitree is not overvalued because it’s overhyped.
It’s undervalued — if its IPO proceeds are used to fund a true AI transition.
Investors should not ask:
“Is Unitree worth $7 billion?”
They should ask:
“Will Unitree use its public market access to build an AI platform — or just sell more legs?”
The answer determines whether it becomes:
- A $10B hardware company — admired, but irrelevant in 10 years
- Or a $100B platform — the foundational layer for embodied AI
The market doesn’t care what you built yesterday.
It cares what you’ll build tomorrow.
Unitree has the cash.
It has the global reach.
It has the credibility.
Now it must prove it has the ambition.
