Company Β· evidence map

Tesla Optimus: capacity-intent curve, not yet S5 proof

Tesla has the clearest public-company capacity-intent signal in humanoid robotics. The public record still supports S4 manufacturing infrastructure rather than S5 commercial economics.

Updated 2026-06-15. Evidence map only; no trade recommendation.

Core Optimus signal board

Capacity intent1M / year

Fremont first-generation Optimus line designed capacity; not current production.

Long-term ambition10M / year

Texas second-generation long-term designed annual capacity; still a preparation signal.

Evidence stageS4

Manufacturing-infrastructure and capacity-intent evidence, not S5 customer economics.

Missing KPIOutput / yield

Accepted units, line yield, uptime, task ROI, revenue and gross margin remain undisclosed.

Optimus Gen 1 / 2 / 3 product progress

This comparison uses Tesla filings for production intent and extracted secondary PDFs for Gen 1/2/3 hardware details. Treat technical specs as a watchlist until source-traced.

Gen 1Prototype stage

Weight: ~73 kg

Speed: 1.7 mph

Manipulation: Early hand / mobility demo

Useful baseline, not deployment proof.

Gen 2Improved mobility + dexterity

Weight: 57 kg

Speed: 5.0 mph

Manipulation: 11-22 DoF hand references in secondary extracts

Secondary extract data; verify against original Tesla material before upgrading.

Gen 3First design framed for mass production

Weight: not confirmed

Speed: not confirmed

Manipulation: 22 DoF hand; forearm + tendon route hypothesis

Hand/actuator details are extracted secondary material, not primary-source proof.

Optimus hand: engineering hypothesis, not S5 proof

Charlie's Optimus PDFs are extracted secondary text. They support a hand/manipulation watchlist, but this page does not treat 22 DoF or forearm-actuator counts as final primary-source proof.

Hand DoF hypothesis22 DoF

Extracted Optimus hand reports describe a move toward 22 DoF, close to the 27-28 DoF human-hand reference.

Actuator packaging~25 / forearm

Gen 3 extract says about 25 micro actuators per forearm including wrist control; treat as secondary until source-traced.

Architecture shiftforearm + tendon

Moving actuators out of the palm can lower distal mass but creates tendon friction, routing and calibration risks.

Commercial gatetask uptime

More DoF only matters if it reduces intervention and survives industrial maintenance cycles.

Sources