Aeva Technologies vs. Cognex: Which Computer Vision Stock Is a Better Buy in 2026?

Aeva Technologies posted FY 2025 revenue of $18.1 million, up 99.4% year over year, but still generated a net loss of about $145.4 million and negative free cash flow of $119.7 million. Cognex reported FY 2025 revenue of roughly $994.4 million, net income of $114.4 million, and free cash flow of $236.8 million, underscoring its more mature and profitable profile. The article favors Cognex over Aeva for 2026, citing lower leverage, stronger cash generation, and a more diversified revenue base despite ongoing competitive and macro risks.

Analysis

The market is still pricing two very different optionalities here: AEVA is a call option on eventual OEM adoption, while CGNX is a compounding franchise with embedded self-funding capacity. The second-order issue for AEVA is not just dilution risk from losses; it is time-to-scale risk, because every quarter spent proving the platform increases the odds that customers either internalize the capability or force price-down adoption cycles. In contrast, CGNX’s advantage is that it can keep investing through a softer industrial tape without tapping capital markets, which tends to widen the competitive gap when smaller vision vendors are forced to cut R&D.

The key catalyst path is asymmetric. AEVA needs a sequence of proof points over the next 6–18 months: design wins, conversion from pilot to volume, and evidence that manufacturing can scale without margin collapse. If that chain breaks, the equity story can rerate sharply lower because the valuation is still anchored to future revenue that must compound very fast to justify the current setup. CGNX does not need a heroic outcome; it only needs mid-single-digit demand stability for earnings power to surprise on the upside because operating leverage and buybacks can do the rest.

The contrarian angle is that the consensus may be underestimating how durable machine vision demand is across non-auto end markets, especially logistics and packaging, where automation budgets are less cyclical than vehicle capex. Conversely, AEVA may be benefiting from a “next-gen lidar” halo that obscures the fact that customer concentration is still a commercialization red flag, not just a growth-stage quirk. The cleaner trade is to own the company with proving power and short the one still in proving mode, rather than paying up for a narrative that needs multiple things to go right simultaneously.

Key Decisions for Investors

• Long CGNX / short AEVAW for 3-6 months: express the spread between cash-generative execution and pre-scale commercialization risk; upside if industrial demand holds and AEVA misses conversion milestones.
• If initiating AEVAW exposure, use only call spreads or small-size event-driven longs into confirmed OEM/program milestones over the next 6-12 months; avoid common equity until revenue inflects with margin discipline.
• Add CGNX on any 5-10% pullback tied to macro/industrial weakness; risk/reward favors a 12-month hold because free cash flow provides downside support and optionality for buybacks/R&D.
• Sell downside on CGNX only after a sharp rally: use put spreads around earnings if implied volatility overshoots, since the balance sheet reduces blowup risk and the stock is more likely to grind than gap lower.
• Avoid being short TSLA as a direct substitute for AEVA weakness; the more likely second-order benefit from lidar skepticism is to established automation names, not necessarily to auto OEMs.