Financial Performance and Outlook

FY26Q1 Financial Performance

• Total revenue: $22.4B, up 16% YoY
• GAAP gross margin: 21.1%, up 478 bps YoY
• Automotive gross margin (GAAP): 21.1%, up from 20.4% in FY25Q4
• Automotive gross margin ex-regulatory credits (non-GAAP): 19.2%, up from 17.9% in FY25Q4
• GAAP operating income: $941M, up 136% YoY; operating margin of 4.2%, up 214 bps YoY
• Adjusted EBITDA: $3.7B, up 30% YoY; adjusted EBITDA margin of 16.4%, up 183 bps YoY
• Non-GAAP diluted EPS: $0.41, up 52% YoY; GAAP diluted EPS: $0.13, up 8% YoY
• Free cash flow: $1.4B

One-Time Items Embedded in Results

• ~$230M warranty true-down benefit in automotive
• ~$250M+ in tariff-related one-time benefits in energy gross margin
• Without these items, underlying profitability improvement would be more limited

Outlook

• FY26 CapEx raised to in excess of $25B (prior guide was $20B+), covering six new production lines plus AI compute and infrastructure
• Management expects negative free cash flow for the remainder of FY26
• Battery pack capacity remains the primary constraint on vehicle production growth; Tesla is actively working to resolve this
• Energy segment deployments are expected to be higher in FY26 vs. FY25 in aggregate, despite the weak Q1; margins expected to compress from competition and tariffs on a normalized basis
• FSD (Supervised) EU-wide approval expected as early as Q2 FY26 following Netherlands approval; China approval targeted for Q3 FY26
• Robotaxi expansion: unsupervised service launched in Dallas and Houston in April; preparations underway for Phoenix, Miami, Orlando, Tampa, and Las Vegas

Business Drivers

Automotive Demand Recovery — but Constrained by Battery Packs

• Demand recovered across EMEA (France and Germany each up over 150% QoQ in deliveries) and APAC (South Korea and Japan growth), with a slight sequential increase in US deliveries as well
• Tesla ended FY26Q1 with the highest Q1 order backlog in over 2 years, with management noting the improvement predated the recent uptick in gas prices — attributing it to more affordable and compelling vehicle trims
• Despite the demand recovery, Tesla's primary production constraint is battery pack capacity, not demand. Management identified this explicitly on both the FY25Q4 and FY26Q1 calls, and noted they are "actively working" to resolve it through 4680 ramps at Gigafactory Texas and Berlin, LFP production ramp in Nevada, and expanded pack capacity in Reno
• Berlin set a record with over 61,000 units produced in FY26Q1, partly enabled by the ramp of in-house 4680 cells in non-structural packs

FSD Adoption Accelerating; Strategy Shifting to FSD-First

• Active FSD subscribers reached approximately 1.28 million globally in FY26Q1, up from 1.10 million in FY25Q4 — the addition of ~180,000 net new subscribers in a single quarter, primarily via subscriptions
• Subscriber churn is declining, which management attributes to improving product quality
• Tesla is repositioning its vehicle sales strategy: the vehicle is now framed as the delivery mechanism, and FSD (Supervised) is the product. This includes a new in-vehicle Self-Driving App for AI4 vehicles where customers can subscribe directly
• FSD received approval in the Netherlands in April 2026, clearing the path for EU-wide approval; China approval is expected by Q3 2026. Management views these as significant unlocks for demand in those markets, where none of the current fleet has access to FSD
• Hardware 3 vehicles (~12% less memory bandwidth than Hardware 4) cannot support unsupervised FSD; Tesla is developing a retrofit program and micro-factory approach to upgrade these vehicles. A V14 distilled version for Hardware 3 is expected end of June 2026

Robotaxi Scaling — Still Early, Still Safe

• Paid Robotaxi miles nearly doubled sequentially in FY26Q1; unsupervised service expanded in Austin and launched in Dallas and Houston in April 2026
• Tesla has had zero safety incidents to date in the unsupervised Robotaxi service
• The primary near-term bottleneck is not safety but operational edge cases: cars getting stuck at unusual intersections, infinite loops around construction, or incorrect drop-off locations. Management described this as a "convenience" problem, not a safety problem
• FSD v14.3 is the current architecture running in Robotaxi markets; v15, a complete software architecture overhaul targeting AI4 hardware, is expected by end of 2026 or early 2027 and is described as a major upgrade
• Unsupervised FSD for personal customer vehicles is expected to roll out gradually starting in Q4 2026, geography by geography, as Tesla validates safety in each market
• FSD miles on the cumulative customer fleet are approaching 10 billion total

Energy Segment: Strong Margin from One-Time Items; Underlying Compression Expected

• Energy revenue declined 12% YoY and 37% sequentially to $2.4B in FY26Q1, with deployments of 8.8 GWh — down 38% sequentially from 14.2 GWh in FY25Q4
• Energy gross margin of over 39.5% in FY26Q1 was driven by more than $250M in one-time tariff benefit recognitions from tariffs paid in prior quarters
• On a normalized basis, management guided for continued energy margin compression from increasing competition and tariff impacts on Chinese LFP cells
• Megapack 3 production at the new Megafactory outside Houston is on track to start later in FY26; this factory will produce the Megapack 3 and Megablock
• Management expects FY26 energy deployments to exceed FY25 in aggregate despite the Q1 weakness

Operating Expenses Rising; AI Investments Ongoing

• Total operating expenses grew 37% YoY to $3.8B in FY26Q1, driven by elevated AI-related R&D (AI5 chip development, Optimus, Cybercab, Semi, Megablock), a full quarter of stock-based compensation under the 2025 CEO award, and SG&A growth
• Management indicated this elevated opex level is expected to continue for the full year
• Cortex 2, the second AI training cluster at Gigafactory Texas, came online in FY26Q1 and began running training workloads; total onsite training compute now exceeds 230,000 H100 equivalents across Cortex 1 and Cortex 2

AI5 Chip Tape-Out and Semiconductor Strategy

• AI5 tape-out was completed in April 2026; the chip was designed to be produced at both TSMC (Arizona) and Samsung (Texas)
• Musk described AI5 as likely to achieve AI inference performance 40x better than AI4 on certain metrics; the chip eliminates the legacy GPU and image signal processor, enabling a half-reticle design
• AI5 is targeted primarily for Optimus and data center deployment; AI4 hardware is sufficient for unsupervised FSD in vehicles, at least for the near term
• Tesla broke ground on its Research Fab at Gigafactory Texas — a ~$3B facility targeting a few thousand wafer starts per month and integrating logic, memory, masking, and packaging under one roof; SpaceX will pursue the larger-scale TerraFab

Optimus Progress; Production S-Curve Expected to Be Slow

• Preparations for the first large-scale Optimus factory (1 million robots/year capacity) in Fremont are beginning in Q2 2026, replacing the Model S/X lines
• A second Optimus factory at Gigafactory Texas is being designed for 10 million robots/year long-term capacity
• Start of production for Optimus 3 is targeted for late July/August 2026, though Musk emphasized the S-curve ramp will be slow given an entirely new supply chain
• Optimus 3 reveal is being delayed deliberately to prevent competitors from conducting frame-by-frame analysis of the design before production begins

One-Offs

• Warranty True-Downs (~$230M): FY26Q1 automotive gross margin benefited from approximately $230M in warranty true-down benefits. This is a one-time item that inflated automotive gross margin above the run-rate level.

• Energy Tariff Recognitions (~$250M+): Energy gross margin of 39.5%+ includes over $250M in tariff-related benefit recognitions from tariffs paid in prior quarters. Management stated that on a normalized basis, energy margins will compress from competition and ongoing tariff impacts. The 39.5% gross margin is not representative of the forward run rate.

• Bitcoin Mark-to-Market: GAAP net income was negatively impacted by mark-to-market losses on Bitcoin holdings, which depreciated approximately 22% QoQ. Bitcoin holdings on the balance sheet declined from $1.0B at FY25Q4 to $786M at FY26Q1 end.

• FX: Unfavorable FX from large intercompany borrowings had a negative impact on reported net income.

• SpaceX Investment: Tesla deployed $2.0B in FY26Q1 to acquire SpaceX equity. This reduces the cash balance meaningfully but is not reflected in operating metrics.

• Model S/X Wind-Down: Final Model S and Model X production ends in early May 2026. The Fremont production line is being dismantled and converted to Optimus production. This removes higher-ASP vehicles from the mix permanently.

• CEO Compensation SBC: FY26Q1 includes a full quarter of SBC under the 2025 CEO Performance Award. One milestone remains deemed probable, contributing to elevated SBC versus the prior year.

Thesis Updates

Automotive Margin Recovery: Real, But Quality Unclear

• Automotive gross margin ex-credits improved to 19.2% in FY26Q1 from 17.9% in FY25Q4, continuing a sequential improvement trend that began in FY25Q2 (15.0%). This supports the bull case that margins are recovering from the FY25Q1 trough.
• However, approximately $230M of the improvement came from warranty true-downs. Stripping this out, the underlying improvement is more modest — an important caveat for investors extrapolating the trend.
• The battery pack capacity constraint on production is a double-edged signal: it confirms demand is exceeding supply (bullish), but it also means Tesla cannot fully capitalize on the demand recovery until the constraint is resolved (near-term limiter).

FSD Adoption Inflecting; Regulatory Unlock Is the Near-Term Catalyst

• The addition of ~180,000 net new FSD subscribers in FY26Q1 — while transitioning to subscription-only — is evidence that the product quality improvement under FSD v14 is converting the installed base at a faster rate. Subscriber churn also declined, which supports the view that retention is improving as the product improves.
• The Netherlands approval in April 2026, with EU-wide approval expected in Q2 2026 and China in Q3, represents a structural unlock. Tesla's entire EMEA and China fleet currently has no access to FSD — if regulatory approvals unlock these markets, penetration has enormous room to grow from its current ~12-13% of the eligible fleet.
• The bear case concern about low adoption despite years of incentives remains partially valid — but the shift toward subscription-only removes the upfront pricing friction that historically discouraged trial.

Robotaxi Still in Early Innings — No Material Revenue Yet

• Tesla's Robotaxi metrics remain limited: paid miles doubled sequentially but were not disclosed in absolute terms; the fleet count per city was not specified. Management acknowledged Robotaxi revenue is not yet meaningful to disclose cost-per-mile or unit economics.
• The expansion to Dallas and Houston in April is a tangible step, but the service remains far from the "half the US population" target management referenced for end of 2026. The operational edge cases described by Musk — cars getting stuck, infinite loops around construction zones, a Waymo-crashing-into-a-bus incident blocking an intersection — illustrate the complexity of scaling beyond carefully validated geofences.
• Musk reiterated that unsupervised FSD for personal customer vehicles is likely available in Q4 2026, which would be the most significant expansion of Robotaxi economics if achieved.

CapEx Escalation Is the Central Bear Risk

• Management raised FY26 CapEx guidance to in excess of $25B, up from the already-elevated $20B+ guide issued at FY25Q4. This is a 3x increase from FY25 actuals of $8.5B.
• Free cash flow will be negative for the remainder of FY26, per management. This is a significant statement from a company that generated $6.2B in free cash flow in FY25 and enters the year with $44.7B in cash.
• The $2B SpaceX equity investment in FY26Q1 (not included in the CapEx figure) adds to total capital outflows. The investment's strategic rationale — Grok as an "orchestra conductor" for Optimus robots — has not been validated and involves a related-party entity.
• Bears will note that the CapEx trajectory keeps escalating: $8.5B in FY25, $20B+ guided in January 2026, now $25B+ in April 2026 — suggesting the ambitions are still expanding faster than the plan is being defined.

Optimus: Further Delayed, But Production Infrastructure Is Real

• Factory preparation for the Fremont Optimus line is beginning in Q2 2026; start of production is targeted for late July/August 2026. This pushes meaningful volume production closer to FY27.
• The decision to delay the Optimus 3 reveal (to prevent competitor reverse engineering) is notable: it implies Tesla believes the design is competitively significant and worth protecting.
• The supply chain challenges remain real: the production S-curve will be slow, unpredictable, and bottlenecked by any of thousands of novel components. Bulls see this as characteristic of every Tesla product launch that eventually scaled; bears see it as a multi-year delay to any material revenue contribution.

Energy: One-Time Boost Masks Structural Compression

• The 39.5% energy gross margin in FY26Q1 looks exceptional but is not the run rate. Management was explicit that normalized margins will compress from here due to competition and tariffs.
• The 38% sequential decline in energy deployments (8.8 GWh vs. 14.2 GWh) reflects the inherently lumpy nature of the business tied to customer deployment timelines. Tesla expects FY26 total deployments to exceed FY25, providing some comfort that the Q1 weakness is timing-related.
• The Megapack 3 launch at Houston later in FY26 is the next product catalyst for the energy segment, but it will not prevent near-term margin compression from Chinese competition.