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

Research library · updated 2026-06-12 · public

Tesla Optimus Filing Evidence Audit v1

Date: 2026-06-12 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe research artifact Visibility: PUBLIC Related files:

  • knowledge/robotics-tesla-optimus-primary-claim-map-v1.md
  • knowledge/robotics-tesla-optimus-evidence-v2.md
  • knowledge/robotics-oem-public-evidence-matrix-v1.md
  • knowledge/robotics-catalyst-map-for-slides-v1.md

Public-safety: yes. No private portfolio data. No trade recommendation.

0. One-line answer

截至 2026-06-12，Tesla Optimus 的公开 primary evidence 可以支持一个清晰但克制的结论：Optimus 已经进入 SEC filing-backed manufacturing-infrastructure / capacity-intent evidence curve；但这些 filings 仍没有披露 actual output、accepted units、factory runtime、intervention rate、external paid customers、robot revenue 或 unit economics。因此 public label 应保持为 S4 infrastructure signal，不是 S5 scaled-commercial-economics proof。🟢/🟠

1. Why this file exists

Codex 已经有 Tesla Optimus primary-source claim map；这份 audit 的增量不是重写 Tesla 深潜，而是做 morning-review 前的核查层：

  1. 哪些句子可以直接从 SEC filing 摘出来。🟢
  2. 哪些 slide/site claim 可以安全保留。🟢/🟠
  3. 哪些压缩时容易越界，必须保留反误读 caveat。🔴/🟠

Verification method: 2026-06-12 在 /home/ubuntu/personal-ai/workspace/finance 用 Python urllib.request 拉取 SEC HTML filing，标准库清洗 HTML 后搜索 Optimus / robotics / production line / mass production / 1 million / 10 million / Digital Optimus / Installed capacity。🟠 method note.

2. Filing audit table

Filing / dateDirect filing-backed evidenceSafe public claimWhat must not be inferredSignal grade
Q1 2025 Form 8-K Exhibit 99.1 / filed 2025-04-22Tesla said it remained on track for builds of Optimus on the Fremont pilot production line in 2025, with wider deployment of bots doing useful work across factories. 🟢Tesla had moved Optimus language beyond pure demo into pilot-production and internal-factory-use intent as of Q1 2025. 🟢No unit count, task list, uptime, intervention rate, productivity, cost saving, or payback disclosed. 🟢/🟠S3/S4
Q2 2025 Form 8-K Exhibit 99.1 / filed 2025-07-23Tesla called Q2 2025 the beginning of a transition from EV / renewable-energy leadership to also becoming a leader in AI, robotics and related services. 🟢Robotics became part of Tesla’s official strategic identity in quarterly reporting. 🟢This does not prove Optimus production, customer demand, revenue, or financial materiality. 🟢/🟠S3
Q3 2025 Form 8-K Exhibit 99.1 / filed 2025-10-22Tesla said first-generation production lines for Optimus were being installed in anticipation of volume production. 🟢Production-line installation is a stronger S4 signal than video/demo evidence. 🟢/🟠“In anticipation” is not actual volume production, utilization, sell-through, or accepted deliveries. 🟢/🟠S4
Q4 2025 Form 8-K Exhibit 99.1 / filed 2026-01-28Tesla said it began installing production lines for Cybercab and fine-tuned its production-primed Optimus design while expanding AI training infrastructure. 🟢Optimus was described as production-primed and connected to Tesla’s physical-AI / manufacturing stack. 🟢/🟠Does not prove completed design validation, field reliability, production yield, or cost curve. 🟢/🟠S4
Q4 2025 Form 8-K Exhibit 99.1 / filed 2026-01-28Tesla said Gen 3 would be its first Optimus design meant for mass production; preparations were underway for the first production line, including supply-chain readiness; start of production planned before end-2026; eventual planned capacity of 1 million robots/year. 🟢This supports mass-production intent, supply-chain-readiness work, and a 1M/year planned-capacity anchor. 🟢Planned capacity and planned start of production are not achieved output or commercial economics. 🟢/🟠S4
Q4 2025 Form 8-K Exhibit 99.1 / filed 2026-01-28Tesla’s infrastructure table listed Robotics / California / Optimus as Construction and stated “Installed capacity ≠ current production rate.” 🟢Tesla itself provides the anti-hype rule: capacity/construction must not be read as current production. 🟢No current production-rate claim is supported by this table. 🟢S4 caution
Q1 2026 Form 8-K Exhibit 99.1 / filed 2026-04-22Tesla said preparations for its first large-scale Optimus factory would begin shortly in Q2; first-generation line designed for 1 million robots/year would replace Model S/X lines in Fremont. 🟢This is the clearest dated Fremont capacity-intent anchor: 1M/year designed line and named site. 🟢Does not prove the line is completed, utilized, producing accepted robots, or profitable. 🟢/🟠S4
Q1 2026 Form 8-K Exhibit 99.1 / filed 2026-04-22Tesla said it was preparing Gigafactory Texas for a second-generation line designed for long-term annual production capacity of 10 million robots. 🟢Texas adds a second-site, long-term 10M/year designed-capacity ambition. 🟢Long-term designed capacity is not a dated production forecast, current capacity, customer demand, revenue, or margin. 🟢/🟠S4
Q1 2026 Form 8-K Exhibit 99.1 / filed 2026-04-22Tesla said first-generation production lines for Optimus were being installed in anticipation of volume production; production rates depend on equipment uptime, component supply, factory-upgrade downtime, regulatory considerations and other factors. 🟢The filing gives execution-risk variables Codex can turn into tracking KPIs: uptime, components, downtime, regulation, tool installation. 🟢The risk variables are not solved just because they are listed. 🟢/🟠S4 caution
Q1 2026 Form 8-K Exhibit 99.1 / filed 2026-04-22Tesla said Digital Optimus represents the next evolution of AI development and an intelligence layer complementing real-world AI for vehicles and humanoid robots; inference compute expansion coincides with Robotaxi and Optimus ramps. 🟢Tesla is linking Optimus to AI / inference / semiconductor / physical-AI infrastructure. 🟢/🟠Does not prove vehicle AI transfers to humanoid manipulation or that vertical integration lowers Optimus unit cost. 🟢/🟠S3/S4

3. Approved public claims for Codex

Codex can safely use these claims if the caveats stay attached:

  1. Tesla Optimus has a filing-backed evidence curve from pilot-line language in Q1 2025 to production-line installation and 1M/10M designed-capacity ambition by Q1 2026. 🟢 as of SEC filings filed 2025-04-22 through 2026-04-22.
  2. The strongest Tesla Optimus evidence is manufacturing infrastructure and capacity intent, not customer-side commercialization proof. 🟠 synthesis from filings.
  3. Fremont 1M/year and Texas 10M/year are designed-capacity / long-term-capacity claims, not actual production rates. 🟢/🟠 Q1 2026 filing plus Tesla capacity caveat.
  4. Tesla’s own filing language says installed capacity is not current production rate and production rates depend on uptime, component supply, downtime, regulatory considerations and other factors. 🟢 Q4 2025 / Q1 2026 filings.
  5. Tesla should sit in the public comparison as “capacity ambition / vertical manufacturing curve,” not as proof that humanoid robots already have customer ROI or full unit economics. 🟠 framework.

4. Do-not-say list

  1. Do not say Tesla is producing 1M Optimus robots/year or 10M Optimus robots/year. 🔴 unsupported.
  2. Do not say Fremont or Texas Optimus lines are completed and utilized unless a later primary filing says so. 🔴 unsupported as of 2026-06-12.
  3. Do not say Tesla has disclosed external paid Optimus customers. 🔴 unsupported in reviewed filings.
  4. Do not say Optimus has proven factory ROI, payback, uptime, intervention rate, safety statistics, or robot gross margin. 🔴 unsupported.
  5. Do not infer any named supplier from Tesla capacity language. 🔴 unsupported; supplier validation requires named customer, order/revenue, value content, and margin evidence.
  6. Do not turn Tesla Optimus evidence into TSLA buy/sell/hold language. 🔴 outside scope / no trade recommendation.

5. Signal vs noise

Signal

  • Repeated SEC-filed Optimus / robotics references across Q1 2025 to Q1 2026. 🟢
  • Production-line installation in anticipation of volume production. 🟢 Q3 2025 / Q1 2026.
  • Gen 3 design meant for mass production and supply-chain readiness language. 🟢 Q4 2025.
  • Named sites and designed-capacity anchors: Fremont 1M/year first-generation line; Texas 10M/year long-term second-generation line. 🟢 Q1 2026.
  • Tesla’s own anti-hype caveat: installed capacity is not current production rate. 🟢 Q4 2025 / Q1 2026.

Noise unless upgraded

  • Viral Optimus videos without task KPI, robot count, or intervention data. 🔴/🟠
  • Capacity screenshots quoted without the “installed capacity ≠ current production rate” caveat. 🟠/🔴
  • Supplier rumors tied to Tesla without filing/customer confirmation. 🔴
  • TAM-only arguments that skip output, yield, reliability, ROI and margin. 🟠

6. What would change our mind

Upgrade from S4 toward S5 only if primary or high-quality customer-side evidence discloses:

  1. Quarterly Optimus output, accepted units, or units deployed internally / externally. 🟢 required.
  2. Production yield, uptime, bottleneck metrics, and cost curve for Fremont or Texas lines. 🟢 required.
  3. Factory deployment KPIs: runtime hours, intervention rate, task success, safety incidents, productivity improvement, cost saving, payback. 🟢 required.
  4. External paid deployment: customer name, robot count, contract value, repeat order, or revenue recognition. 🟢 required.
  5. Financial materiality: robot revenue, gross margin, capex line, segment disclosure, or service/support cost. 🟢 required.

7. Slide/site compression

Slide title: Tesla Optimus: filing evidence is stronger, but still not S5

Three-column layout:

Proven by filingsWhy it mattersMissing before S5
Fremont pilot-line builds in 2025; production lines installed; Gen 3 meant for mass production; Fremont 1M/year designed line; Texas 10M/year long-term designed line. 🟢Tesla is treating Optimus as a manufacturing-scale physical-AI project, not only a demo. 🟠Actual output, yield, deployed units, runtime, intervention rate, external paid customer, robot revenue, gross margin, ROI/payback. 🟢/🟠

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

8. Common misconceptions

Misconception 1: “1M / 10M means current production.”

Correction: Tesla’s filings use designed capacity / long-term annual capacity language, and its own capacity-table caveat says installed capacity is not current production rate. 🟢

Misconception 2: “Tesla’s internal factory path is equivalent to external commercialization.”

Correction: internal deployment can be the right first testbed, but S5 requires operating data: robot count, hours, task success, intervention rate, safety, productivity, payback and repeatability. 🟠

Misconception 3: “Tesla capacity ambition validates all humanoid suppliers.”

Correction: supplier value capture still needs confirmed relationship, order/revenue, value content, ASP, gross margin, architecture fit and replacement risk. 🟠

9. Think Deeper questions

  1. If Tesla first deploys Optimus internally, should the key KPI be robot revenue or factory productivity / payback?
  2. Which evidence upgrades the thesis first: actual output, accepted units, runtime hours, intervention rate, or customer ROI?
  3. If Tesla reaches high designed capacity but keeps cost/reliability data private, how much should public evidence frameworks infer from capacity alone?
  4. If Tesla vertically integrates core modules, where can external suppliers still capture defensible value?

10. Source list

11. 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 actual production.
  • Claims that internal factory intent equals external customer economics.