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

Research library · updated 2026-06-18 · public

Tesla Optimus evidence denominator ladder — produced vs accepted vs deployed vs productive vs economic

Date: 2026-06-18 As-of: 2026-06-18 20:09 UTC Owner: Hugo / Genius Team Research owner: Finance / Charlie AGT-002 Status: SYNTHESIS_CANDIDATE Visibility: PUBLIC Output target: morning review / Tesla Optimus slide guardrail / /robotics/tesla-optimus/ evidence-denominator module

Related files:

  • knowledge/robotics-research-harness.md
  • knowledge/robotics-tesla-optimus-sec-filing-excerpt-appendix-2026-06-16.md
  • knowledge/robotics-tesla-optimus-source-to-slide-claim-control-2026-06-17.md
  • knowledge/robotics-tesla-optimus-s5-trigger-dashboard-2026-06-18.md

Public-safety flag: PUBLIC if used as written. Excludes Hugo private portfolio data, watchlist sizing, private channel checks, supplier rumors, paid-report excerpts, tax/legal context, and TSLA buy / sell / hold language.

0. One-line answer

截至 2026-06-18 20:09 UTC，Tesla Optimus 的公开证据已经足以证明一个 S4 capacity-intent / manufacturing-infrastructure curve：2025-04-22 到 2026-04-22 的 Tesla SEC 8-K Exhibit 99.1 序列反复出现 Optimus / robotics / production-line / mass-production / 1M / 10M designed-capacity anchors。🟢 但这些 filings 仍没有披露 produced units、accepted units、deployed units、productive hours、intervention rate、factory ROI/payback、external paid customers、Optimus revenue、gross margin 或 service burden；因此还不能写成 S5 scaled-commercial-economics proof。🟢 Tesla filings; 🟠 Charlie evidence-denominator synthesis.

1. Why this artifact exists

已有 Tesla 文件已经回答 “filings 说了什么”。本文件只补一个更适合晨会判断的分母框架：

Tesla Optimus 的下一步证据不能只问“产能多大”，而要问披露的是 produced、accepted、deployed、productive 还是 economic denominator。

这五个分母决定证据等级：

  1. Produced denominator: 造了多少。
  2. Accepted denominator: 合格/验收了多少。
  3. Deployed denominator: 放到哪个场景、多少台、干什么任务。
  4. Productive denominator: 工作了多少小时、干预多少、故障多少、产出多少。
  5. Economic denominator: 节省/创造多少价值，成本、服务、毛利和回收期如何。

2. Fresh primary-source verification

SEC pages were rechecked from /home/ubuntu/personal-ai/workspace/finance on 2026-06-18 20:09 UTC using Python urllib.request, SEC HTML stripping, and keyword search. 🟠 verification method; source pages are 🟢 primary.

Filing anchorSEC status on 2026-06-18 20:09 UTCRelevant keyword hitsSafe evidence role
Q1 2025 Form 8-K Exhibit 99.1, filed 2025-04-22HTTP 200Optimus; robotics; production lines; volume production; Installed capacityFremont pilot production line / internal factory-use intent. 🟢
Q2 2025 Form 8-K Exhibit 99.1, filed 2025-07-23HTTP 200robotics; volume production; Installed capacityRobotics enters Tesla official strategic reporting identity; not Optimus production proof. 🟢
Q3 2025 Form 8-K Exhibit 99.1, filed 2025-10-22HTTP 200Optimus; robotics; production lines; volume production; Installed capacityFirst-generation Optimus production lines being installed in anticipation of volume production. 🟢
Q4 2025 Form 8-K Exhibit 99.1, filed 2026-01-28HTTP 200Optimus; robotics; production lines; volume production; mass production; Installed capacityProduction-primed Optimus design, Gen 3 mass-production intent, first-line preparation, supply-chain readiness, eventual 1M/year planned-capacity anchor. 🟢
Q1 2026 Form 8-K Exhibit 99.1, filed 2026-04-22HTTP 200Optimus; robotics; production lines; mass production; 1 million; 10 million; Digital OptimusFremont first-generation line designed for 1M robots/year and Texas second-generation line designed for long-term 10M robots/year. 🟢

Important verification caveat: keyword hits show primary-source presence, not economic proof. The same filings include Tesla’s capacity-table caution that installed capacity is not current production rate. 🟢

3. Evidence denominator ladder

DenominatorWhat the number would answerExample of acceptable disclosureCurrent status as of 2026-06-18Evidence grade / stage impact
0. Designed capacityHow much capacity Tesla intends to build or design for.Fremont line designed for 1M robots/year; Texas line designed for long-term 10M robots/year.Present in Q1 2026 filing.🟢 Primary; S4 capacity-intent evidence. Not production proof.
1. Produced unitsHow many Optimus units were built in a period.“X Optimus units produced in Qx 2026,” with date and site.Missing in reviewed filings.Would upgrade from capacity-intent to production-proof, but still not S5 alone. 🟢 needed.
2. Accepted unitsHow many units passed EOL / quality / safety acceptance.Produced count plus FPY / EOL yield / accepted unit count / rework rate.Missing.More valuable than produced units because it separates output from usable robots. 🟢 needed.
3. Deployed unitsWhere robots are actually working and doing what task.Named factory, robot count, task, shift pattern, operating envelope, start date.Missing beyond broad internal factory-use intent.Starts deployment-proof path; still needs productivity and intervention. 🟢/🟡 needed.
4. Productive hoursWhether deployed robots are doing useful work reliably.Fleet runtime hours, uptime %, intervention rate, reset/failure rate, safety incidents, task success.Missing.Strong S5 candidate only when paired with accepted/deployed denominator. 🟢/🟡 needed.
5. Economic outputWhether Optimus creates measurable financial value.Labor-hour substitution, throughput improvement, quality improvement, cost saving, ROI/payback months, robot revenue, gross margin, support/warranty burden.Missing.S5 economics proof if disclosed with denominator; most thesis-changing. 🟢/🟡 needed.

4. What can be said publicly now

Allowed wording:

  • “Tesla Optimus has filing-backed capacity-intent evidence, including a Fremont first-generation line designed for 1M robots/year and a Texas second-generation line designed for long-term 10M robots/year, as disclosed in Tesla Q1 2026 Form 8-K Exhibit 99.1 filed 2026-04-22.” 🟢
  • “This is S4 manufacturing-infrastructure evidence, not S5 commercial-economics proof.” 🟠 synthesis from 🟢 filings.
  • “The next proof is not another capacity number; it is the denominator chain from produced units to accepted units to deployed productive hours to economics.” 🟠
  • “Tesla’s own capacity caveat says installed capacity is not current production rate.” 🟢

Forbidden wording unless future primary evidence changes:

  • “Tesla is producing 1M Optimus robots/year.” 🔴 unsupported by reviewed filings.
  • “Tesla has 10M/year Optimus capacity today.” 🔴 unsupported by reviewed filings.
  • “Optimus has proven factory ROI/payback.” 🔴 missing denominator and economics.
  • “Optimus has external paid customers.” 🔴 missing in reviewed filings.
  • “Supplier winners are confirmed by Tesla capacity language.” 🔴 capacity ambition is not named supplier/customer proof.

5. Signal vs noise

Signal

  • Five SEC-filed Tesla quarterly updates from 2025-04-22 to 2026-04-22 are reachable as of 2026-06-18 20:09 UTC and contain Optimus / robotics / production-line / capacity anchors. 🟢
  • Q3 2025 and Q1 2026 use production-line installation language for Optimus. 🟢
  • Q4 2025 frames Gen 3 as the first design meant for mass production and references first-line preparation plus supply-chain readiness. 🟢
  • Q1 2026 provides named-site and quantified designed-capacity anchors: Fremont 1M/year designed line; Texas long-term 10M/year designed line. 🟢
  • Tesla’s own capacity-table caveat prevents equating capacity with current production rate. 🟢

Noise unless upgraded

  • Demo video without robot count, task denominator, runtime, intervention, failure/reset, safety or productivity metrics. 🔴/🟠
  • Produced-unit rumor without primary source or without accepted/deployed denominator. 🔴/🟠
  • Supplier rumor based on capacity ambition without named customer, order value, revenue recognition, teardown, or official relationship. 🔴
  • Customer/logo claim without contract value, robot count, delivery schedule, repeat order, task KPI, ROI/payback or support burden. 🟠
  • “AI stack / Digital Optimus” language treated as proof of manipulation economics. 🟠

6. Morning-review use: five denominator cards

Title: Tesla Optimus: capacity is the starting denominator, not the finish line

Card 1 — Designed capacity already exists 🟢

  • Fremont first-generation line designed for 1M robots/year.
  • Texas second-generation line designed for long-term 10M robots/year.
  • Stage: S4 capacity-intent / manufacturing-infrastructure.

Card 2 — Produced is still missing 🔴

  • Need period-by-period produced-unit count.
  • Better if split by Fremont / Texas, Gen 2 / Gen 3, engineering build / production build.

Card 3 — Accepted beats produced 🟠

  • Need FPY / EOL yield / accepted units / rework / component bottlenecks.
  • High output without acceptance can be manufacturing churn.

Card 4 — Deployed + productive is the first real S5 bridge 🟠

  • Need robot count by factory/task, runtime hours, uptime, intervention, safety and productivity.
  • Internal factory deployment can be economically relevant if denominator and payback are disclosed.

Card 5 — Economics closes the loop 🟢/🟡 needed

  • Need labor-hour substitution, throughput/quality improvement, cost saving, ROI/payback, revenue, gross margin and service burden.
  • Without this, Tesla remains S4 even with strong capacity ambition.

Required footer: Evidence map only. No TSLA trade recommendation. Designed capacity is not current production. Produced units are not accepted deployments. Internal factory intent is not external paid-customer economics.

7. What would change our mind

Upgrade toward S5 if Tesla or high-quality customer-side evidence discloses at least 3 of the following together:

  1. Produced units by quarter and site. 🟢
  2. Accepted units / FPY / EOL yield / rework or bottleneck data. 🟢
  3. Deployed robot count by factory, task and shift pattern. 🟢
  4. Fleet runtime hours plus uptime / intervention / reset / safety incident distribution. 🟢/🟡
  5. Factory productivity improvement, labor-hour substitution, cost saving, ROI/payback months. 🟢/🟡
  6. External paid customer with robot count, contract value, delivery schedule, repeat order or revenue recognition. 🟢
  7. Optimus revenue, gross margin, warranty/service/support burden, or segment-level financial disclosure. 🟢

Downgrade or hold at S4 if:

  • Capacity ambition grows but produced/accepted/deployed/productive/economic denominators stay absent. 🟠
  • Production timeline slips without replacement metric. 🟢/🟡 depending on source.
  • Output appears but no yield, acceptance, runtime or economics are disclosed. 🟠
  • Market narratives infer supplier/customer winners without primary evidence. 🔴/🟡.

8. Common misconceptions

Misconception 1: “1M / 10M capacity is the same as production.”

Correction: 1M and 10M are designed / long-term designed capacity anchors in reviewed filings, not current production rates. Tesla’s own capacity caveat says installed capacity is not current production rate. 🟢

Misconception 2: “Produced units would automatically prove commercialization.”

Correction: produced units only answer build volume. Commercialization still needs accepted units, deployment denominator, runtime/intervention, ROI/payback, revenue, margin and service burden. 🟠

Misconception 3: “Internal deployment is irrelevant because it is not external sales.”

Correction: internal deployment could be the first S5 path if Tesla proves productivity and payback inside its own factories. But without robot count, runtime/intervention and ROI, it remains intent/deployment evidence rather than economics proof. 🟠

Misconception 4: “Tesla capacity validates supplier winners.”

Correction: supplier value capture needs named relationship, order/revenue, value content, margin, architecture durability and replacement risk. Tesla capacity language alone does not confirm any supplier. 🟠

9. Think Deeper questions

  1. Would 1,000 accepted robots with 1,000,000 runtime hours and low intervention rate matter more than 10,000 produced robots with no yield or task data?
  2. If Tesla first proves Optimus through internal productivity, should the public KPI be robot revenue or productivity value per deployed robot?
  3. Which denominator should trigger a supplier thesis update: Tesla output, Tesla accepted units, teardown evidence, or supplier filing revenue?
  4. If future filings show produced units but no economics, should the stage move to S4+ production proof rather than S5 return proof?
  5. What is the minimum denominator package that would make Optimus comparable to an EV ramp rather than a demo-to-capacity story?

10. Source list

11. Public-safe / internal-only boundary

Public-safe:

  • The denominator ladder, allowed/forbidden wording, signal/noise list, morning-review cards, misconception section, Think Deeper prompts and source list above.

Keep internal-only / exclude:

  • Hugo private portfolio data, watchlist sizing, private channel checks, supplier rumors, paid-report excerpts, tax/legal context, TSLA buy/sell/hold language, or inferred supplier/customer winners not supported by primary or high-quality customer-side evidence.