Investment ResearchTopic research, evidence curves, and cycle signals
HomeSign inEN

Research library · updated 2026-06-14 · public

Tesla Optimus Public Evidence Pack v2

Date: 2026-06-14 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe research artifact Visibility: PUBLIC Target use: slide 11, /robotics/companies, /robotics/signals, or Tesla evidence box inside morning review

Related files:

  • knowledge/robotics-research-harness.md
  • knowledge/robotics-tesla-optimus-primary-claim-map-v1.md
  • knowledge/robotics-tesla-optimus-filing-evidence-audit-v1.md
  • knowledge/robotics-tesla-optimus-codex-packaging-brief-v1.md
  • knowledge/robotics-tesla-figure-unitree-public-evidence-ledger-v1.md

Public-safety: yes. No Hugo private portfolio data. No TSLA buy / sell / hold language. No external publishing action.

0. One-line answer

截至 2026-06-14，Tesla Optimus 的公开证据最适合被写成“manufacturing-infrastructure / capacity-intent curve”：Tesla SEC-filed quarterly updates 已经给出 Fremont pilot-line build intent、production-line installation、Gen 3 mass-production-oriented design、Fremont 1M/year designed line、Texas 10M/year long-term designed line等 S4 证据；但截至已复核 filings，仍没有 actual output、accepted units、yield、runtime、intervention rate、external paid customers、robot revenue、gross margin 或 ROI/payback。🟢/🟠

1. Why this artifact exists

已有 Tesla Optimus claim map / filing audit / Codex packaging brief。这个 v2 的增量是：

  1. 重新在线验证 SEC filing URL 可访问，并确认关键关键词在原始 HTML 中出现。🟢 as of 2026-06-14.
  2. 把 Tesla public evidence 压缩成 morning-review / slide-ready 的“what is proven / what is not proven / what to track next”。🟠
  3. 给 Codex 一个可直接包装的 public-safe evidence box，避免把 designed capacity 误写成 current production。🟠

Verification run: 2026-06-14 在 /home/ubuntu/personal-ai/workspace/finance 用 Python urllib.request 拉取 5 个 SEC HTML filing，并计数 Optimus / Robotics / production lines / 1 million / 10 million / Digital Optimus / Installed capacity。🟠 method note.

2. Verification log

FilingSEC URL status on 2026-06-14Keyword checkSource grade
Q1 2025 Form 8-K Exhibit 99.1, filed 2025-04-22OK, 49,106 HTML chars fetchedOptimus 1; Robotics 1; production lines 3; Installed capacity 1🟢 primary
Q2 2025 Form 8-K Exhibit 99.1, filed 2025-07-23OK, 48,477 HTML chars fetchedRobotics 3; Installed capacity 1; no Optimus hit in keyword count🟢 primary
Q3 2025 Form 8-K Exhibit 99.1, filed 2025-10-22OK, 47,988 HTML chars fetchedOptimus 1; Robotics 2; production lines 2; Installed capacity 1🟢 primary
Q4 2025 Form 8-K Exhibit 99.1, filed 2026-01-28OK, 61,107 HTML chars fetchedOptimus 5; Robotics 4; production lines 4; 1 million 3; Installed capacity 2🟢 primary
Q1 2026 Form 8-K Exhibit 99.1, filed 2026-04-22OK, 49,885 HTML chars fetchedOptimus 7; Robotics 5; production lines 2; 1 million 1; 10 million 1; Digital Optimus 1🟢 primary

Important limitation: keyword presence verifies the filings were reachable and contain the expected evidence anchors; the already-existing claim map / audit files remain the detailed snippet-level extraction layer. 🟠

3. Evidence curve: what is proven vs not proven

Evidence stepSafe public claimWhat it provesWhat it does not proveSignal grade
Q1 2025 Fremont pilot-line languageTesla disclosed it remained on track for Optimus builds on a Fremont pilot production line in 2025 and wider deployment of bots doing useful work across factories.Optimus had moved beyond pure demo language into pilot-production / internal-use intent. 🟢 as of 2025-04-22No unit count, task list, uptime, intervention rate, productivity, cost saving, payback, or safety data. 🟢/🟠S3/S4
Q2 2025 robotics strategic framingTesla described its transition toward AI, robotics and related services.Robotics was part of Tesla’s official strategic identity in quarterly reporting. 🟢 as of 2025-07-23Does not prove Optimus production output, customer demand, revenue, or financial materiality. 🟢/🟠S3
Q3 2025 production-line installationTesla disclosed first-generation production lines for Optimus were being installed in anticipation of volume production.Production-line installation is materially stronger than viral demo evidence. 🟢 as of 2025-10-22“In anticipation” is not actual volume output, utilization, sell-through, or accepted units. 🟢/🟠S4
Q4 2025 Gen 3 / mass-production designTesla described Gen 3 as its first Optimus design meant for mass production, with first-line preparations and supply-chain readiness.Product / supply-chain work was being framed for manufacturability. 🟢 as of 2026-01-28Does not prove completed validation, yield, reliability, cost curve, or margin. 🟢/🟠S4
Q4 2025 1M/year planned capacity anchorTesla disclosed eventual planned capacity of 1M robots/year for the first production line and warned installed capacity is not current production rate.Capacity ambition became quantified, while Tesla’s own caveat gives the anti-hype rule. 🟢 as of 2026-01-28Planned capacity is not achieved output or current production rate. 🟢/🟠S4 caution
Q1 2026 Fremont 1M/year designed lineTesla said preparations for its first large-scale Optimus factory would begin shortly in Q2; first-generation line designed for 1M robots/year would replace Model S/X lines in Fremont.Strongest dated Fremont capacity-intent anchor: named site + 1M/year designed line. 🟢 as of 2026-04-22Does not prove line completion, utilization, accepted robots, customer ROI, or profitability. 🟢/🟠S4
Q1 2026 Texas 10M/year long-term lineTesla said Gigafactory Texas was being prepared for a second-generation Optimus line designed for long-term annual capacity of 10M robots.Adds second-site, long-term scale ambition. 🟢 as of 2026-04-22Long-term designed capacity is not a dated output forecast, customer order, revenue, or gross margin. 🟢/🟠S4
Q1 2026 Digital Optimus / inference languageTesla linked Digital Optimus, humanoid robots, Robotaxi ramps, AI inference compute and semiconductor / fabrication strategy.Optimus is tied to Tesla’s AI / inference / manufacturing stack. 🟢/🟠 as of 2026-04-22Does not prove vehicle AI transfers to humanoid manipulation, deployment safety, or robot unit cost. 🟢/🟠S3/S4

4. Public-safe Tesla Optimus evidence box

Tesla Optimus: capacity ambition is evidence, not proof

What filings prove 🟢

  • 2025 Fremont pilot-line build intent.
  • First-generation production-line installation in anticipation of volume production.
  • Gen 3 design meant for mass production.
  • Fremont first-generation line designed for 1M robots/year.
  • Texas second-generation line designed for long-term annual capacity of 10M robots.
  • Tesla’s own caveat: installed capacity is not current production rate; production rates depend on uptime, component supply, factory-upgrade downtime, regulatory considerations and other factors.

Why it matters 🟠

  • Tesla Optimus is no longer just a demo evidence curve; it has entered an official manufacturing-infrastructure evidence curve.
  • Named sites, line replacement, production-line installation, supply-chain readiness and quantified designed capacity are stronger evidence units than social-media videos.
  • This is a material S4 infrastructure signal for the robotics field guide.

What is missing before S5 🟢/🟠

  • Quarterly Optimus output or accepted units.
  • Production yield, bottleneck and cost-curve data.
  • Robot count deployed in Tesla factories.
  • Runtime, uptime, intervention rate, task success and safety incidents.
  • Internal productivity improvement, labor substitution, ROI/payback or operating-cost reduction.
  • External paid customer, contract value, repeat order, pricing or revenue recognition.
  • Robot revenue, gross margin, warranty/service cost, capex attribution or segment disclosure.

Footer: Evidence map only. No TSLA trade recommendation. Designed capacity ≠ current production rate.

5. Signal vs noise

Signal

  • SEC-filed repeated robotics / Optimus references across Q1 2025 through Q1 2026. 🟢 as of reviewed filings through 2026-04-22.
  • Production-line installation / construction / tool-installation language. 🟢
  • Gen 3 design meant for mass production and supply-chain readiness language. 🟢
  • Fremont 1M/year and Texas 10M/year designed-capacity anchors. 🟢
  • Tesla’s own anti-hype caveat around installed capacity vs current production rate and execution variables. 🟢

Noise unless upgraded

  • Viral Optimus demo videos without robot count, task KPI, runtime or intervention data. 🟠/🔴
  • Any 1M / 10M screenshot without “designed / long-term capacity” caveat. 🔴 if presented as current production.
  • Supplier rumors tied to Optimus without named customer, order, revenue, teardown or filing-backed exposure. 🔴
  • TAM-only Optimus valuation arguments without output, yield, deployment economics and margin. 🟠
  • Internal-use language without measurable factory operating KPI. 🟠

6. What would change our mind

Upgrade from S4 toward S5 only if primary or high-quality customer-side evidence discloses at least several of the following:

  1. Quarterly Optimus output, accepted units, or units deployed internally / externally. 🟢 required.
  2. Production yield, bottleneck, tool uptime, component supply, or cost curve. 🟢 required.
  3. Factory operating KPI: runtime hours, uptime, intervention rate, task success, safety incidents. 🟢 required.
  4. Quantified productivity improvement, labor substitution, ROI/payback, or operating-cost reduction. 🟢/🟡 preferred.
  5. External paid deployment: customer name, robot count, contract value, repeat order, pricing, or revenue recognition. 🟢 required for strongest commercial proof.
  6. Financial materiality: robot revenue, gross margin, warranty/service cost, capex attribution, or segment disclosure. 🟢 filing preferred.

Downgrade if:

  • Tesla keeps disclosing capacity language for multiple quarters without output / deployment KPI. 🟠 monitoring rule.
  • Start-of-production timing slips without primary-source explanation. 🟢/🟠 source-dependent.
  • Capacity progresses but no cost, reliability, safety or payback evidence appears. 🟠 framework.

7. Common misconceptions

Misconception 1: “1M / 10M means Tesla is already producing at those rates.”

Correction: Tesla’s filings use designed / long-term capacity language, and existing Tesla capacity-table caveats say installed capacity is not current production rate. 🟢

Misconception 2: “Internal Tesla factory use equals external commercialization.”

Correction: internal factory use can be a strong testbed, but S5 requires quantified productivity, uptime, intervention, safety, repeatability, ROI/payback and eventually customer/revenue evidence. 🟠

Misconception 3: “Vehicle AI automatically solves humanoid robotics.”

Correction: shared AI / inference infrastructure may help, but humanoid manipulation adds balance, dexterity, contact-rich control, tool use, human-proximity safety and long-horizon physical reliability. 🟠

Misconception 4: “Optimus capacity validates all humanoid suppliers.”

Correction: supplier value capture requires confirmed relationship, order/revenue, value content, volume, ASP, gross margin, architecture fit and replacement-risk analysis; Tesla may vertically integrate or redesign modules. 🟠

8. Think Deeper questions

  1. If Optimus first scales inside Tesla factories, should investors measure robot revenue first, or factory productivity / payback first?
  2. Which KPI would upgrade the thesis faster: accepted units, production yield, runtime hours, intervention rate, or ROI/payback?
  3. If Tesla reaches large designed capacity but keeps cost and reliability data private, how much should public evidence frameworks infer from capacity alone?
  4. If Tesla vertically integrates core robot modules, where can external suppliers still capture defensible value: precision parts, materials, batteries, sensors, testing equipment, or none?

9. Source list

10. Public-safe flag

Public-safe: yes.

Exclude:

  • Hugo private portfolio data.
  • TSLA buy / sell / hold language.
  • Supplier rumors or inferred named suppliers.
  • Claims that designed capacity equals current production.
  • Claims that internal factory intent equals external customer economics.
  • Paid-report excerpts or private channel information.