News · Funding
Humanoid funding in 2026 — production capacity, not just valuations
Source: Figure AI
Lead strip: Status: capital markets hot / field fleets uneven · Availability: varies by platform · All-in cues: valuations ≠ list prices · Scoring: methodology
Evidence brief only — not a sold ranking. Funding ≠ fleet. Designed factory capacity ≠ robots earning paychecks.
2025–2026 capital flooded humanoid robotics. Valuations jumped while fewer programs published named deployments, factory claims, or order-book milestones. This brief maps raises to production capacity and paid-deployment evidence — capital markets separate from field fleets. Criteria: methodology.
Robot strips (context, not a trophy board)
- Digit · Agility Robotics · Status: Digit 4 field / Digit 5 announced · Use: warehouse logistics · Evidence: named sites + hours/totes · → /robots/digit/
- Figure 03 · Figure AI · Status: Pilot (BMW Spartanburg sequencing) · Use: factory logistics · Evidence: BMW partner + BotQ claims · → /robots/figure-03/
- Phoenix · Sanctuary AI · Status: Research / pilot · Use: general-purpose workplace · Evidence: capital + short pilots; fleet hours — · → /robots/phoenix/
- Optimus · Tesla · Status: Research / production-ramp · Use: factory R&D / Academy · Evidence: Tesla IR / CapEx; external buy page — · → /robots/optimus/
The rule that matters: funding ≠ fleet
A raise documents investor belief and runway. It does not by itself document:
- Multi-site paid operational hours
- Third-party throughput KPIs
- A public list price or RaaS rate card
- That “designed capacity” is actually shipping
Score capital under continuity. Score fleets under paid deployments / third-party record. Mixing them produces trophy narratives our methodology rejects.
Capital map (selected Western programs)
| Program | Capital signal (public) | Production / capacity claim | Paid-deployment evidence class |
|---|---|---|---|
| Figure AI | Series B $675M (2024, $2.6B); Series C >$1B at $39B post-money (Sep 2025) | BotQ designed up to ~12,000/yr; 350+ Figure 03 units claimed (Apr 2026 vendor post) | BMW partner sequencing for Figure 03; Figure 02 history separate |
| Agility / Digit | SPAC path ~$2.5B pre-money framing; $300M+ Digit 5 multi-year orders (May 2026, milestone-conditioned) | RoboFab designed up to 10,000 Digits/yr | Digit 4: >65k hours, 100k+ GXO totes — Digit |
| Apptronik / Apollo | Series A ~$415M (2025) + ~$520M extension (Feb 2026); valuation headlines ~$5–5.5B | Jabil manufacturing partnership; mass-production prep language | GXO / Mercedes / Jabil pilot-zone language; published hours — |
| Sanctuary / Phoenix | >$140M total investment claim incl. BDC/InBC + prior SIF (2024 company) | Humanoid + Physical AI software expansion (2026) | Short pilots / Research — Phoenix; fleet hours — |
| Tesla / Optimus | Internal CapEx: Tesla raised 2026 spend guidance to >$25B company-wide (AI, robotics, chips — Optimus share —) | Ambition language: ~1M/yr Fremont-scale / 10M long-term Texas narratives in secondary IR coverage | Research / production-ramp; external buy page — — Optimus |
Sources for the table cells are primary company releases and major wire coverage cited below. Where a number is vendor capacity design or IR ambition, label it as such — not as 2026 shipped fleet.
What the big raises actually buy
Figure: valuation to BotQ and BMW
Figure’s September 2025 Series C (>$1B committed, $39B post-money) sits on top of the 2024 $675M commercialization round (Figure Series C; Reuters). Capital funded AI training, headcount, and manufacturing scale.
Capacity claims: Figure’s April 2026 production post claims BotQ delivered over 350 third-generation robots and moved toward 1 robot/hour cadence with yield metrics (Figure production ramp). Designed annual capacity around 12,000 appears in industry summaries of BotQ ambition.
Fleet claims that score: BMW Group’s June 2026 statement that Figure 03 starts logistics sequencing at Spartanburg after a completed Figure 02 body-shop pilot (BMW Group). Figure 02 ≠ Figure 03 — do not assign the 30,000-vehicle body-shop history to Figure 03. Product page: Figure 03.
Public catalog buy path for Figure 03: —.
Agility: orders and RoboFab vs Digit 4 hours
Agility’s capital-markets story (SPAC combination framing at roughly $2.5B) runs beside a rarer asset: published Digit 4 field depth. The Digit 5 unveil cites >65,000 Digit 4 hours and 100,000+ totes at GXO Flowery Branch, plus >$300M multi-year Digit 5 orders subject to milestones as of May 2026 (Digit 5 unveil).
RoboFab designed capacity up to 10,000/year is a factory design figure — not proof that 2026 output matches the nameplate. Digit 4 ≠ Digit 5: hours stay on Digit 4 until Digit 5 early access (H1 2027 framing) ships. See Digit.
Apptronik: capital ahead of published hours
Apptronik’s ~$520M February 2026 extension (Google, Mercedes-Benz, and others; valuation headlines ~$5–5.5B) targets Apollo development and production ramp (Reuters). Partner-zone pilots (Mercedes-Benz, GXO, Jabil) are real; published Apollo warehouse hours remain —. Capital ≠ Digit-class tote ledger.
Sanctuary Phoenix: Canadian capital, Research status
Sanctuary’s >$140M total investment claim (BDC/InBC, prior SIF, strategic corporates) documents ecosystem support, not a multi-site fleet certificate (Sanctuary). Page: Phoenix. Fleet hours / list price: —.
Optimus: CapEx ambition without an external buy page
Tesla’s >$25B 2026 CapEx guidance is company-wide AI/robotics/chip spend (TechCrunch). Optimus share: —. Million-unit capacity narratives remain ambition / IR talk, not a public buy channel. Page: Optimus.
How to read capacity numbers
| Phrase | Safe reading | Unsafe reading |
|---|---|---|
| “Factory designed for N/year” | CapEx and tooling intent | “Shipping N this year” |
| “350+ robots delivered” (vendor) | Hardware volume for data + pilots | All units in customer production |
| “$300M+ multi-year orders” | Contracted potential if milestones hit | Cash revenue today |
| “$39B valuation” | Capital-markets clearing price | Field ranking or best robot |
| Partner pilot at named plant | Third-party willingness to trial | Open buy path for all buyers |
Industry trackers (e.g. Dealroom-style tallies of multi-billion humanoid venture dollars through mid-2026) are useful for market temperature. They are not deployment scoreboards.
Practical takeaway for buyers and analysts
- Split ledgers: valuation / raise size ≠ paid fleet hours ≠ designed capacity.
- Ask for the factory and the floor: nameplate robots/year, then nameplate customer hours and intervention rates.
- Prefer primary partner statements (BMW for Figure 03; GXO/Agility for Digit) over valuation headlines.
- Cross-read Digit, Figure 03, Phoenix, Optimus, and methodology.
FAQ
Does a $39B valuation mean Figure leads deployments?
No. Valuation is a capital-markets clearing price. Deployment evidence for Figure 03 is scored from BMW partner materials and vendor manufacturing claims — separately from Series C size. Funding ≠ fleet.
Is designed capacity the same as production output?
No. RoboFab’s 10,000/year or BotQ’s ~12,000/year design figures are capacity intent. Shipped units, yields sustained over a full year, and customer-installed fleets are different metrics.
Which programs show the densest paid-deployment public record?
Digit (Digit 4 hours/totes) has the densest published warehouse narrative here. Figure 03 has a BMW partner factory sequencing trail. Phoenix and Optimus are lighter on public fleet KPIs. Evidence classification — not a sold ranking.
Should CapEx guidance count like venture funding?
Treat Tesla’s CapEx raise as internal capital allocation for AI/robotics/chips — useful context for Optimus, but not equivalent to a priced venture round or a customer RaaS receipt. Optimus share of the envelope remains — unless Tesla breaks it out.
Sources
- Figure AI: Series C (Sep 2025); Ramping Figure 03 Production (29 Apr 2026)
- Reuters: Figure $39B valuation
- BMW Group: Figure 03 Spartanburg (25 Jun 2026)
- Agility Robotics: Digit 5 unveil (15 Sep 2026)
- Reuters: Apptronik $520M (11 Feb 2026)
- Sanctuary AI: BDC / InBC financing (3 Jul 2024)
- TechCrunch: Tesla CapEx >$25B (22 Apr 2026)
- Robotscomparison: Digit, Figure 03, Phoenix, Optimus, methodology
Platform-specific list prices, Optimus CapEx share, Apollo fleet hours, and Phoenix scaled third-party hours: —.