Approximately 20 GWh of US-made cell capacity has publicly verifiable evidence of accepted commercial shipments as of August 2026. This is the figure the registry exists to track.
It sits against roughly 100 GWh of nameplate capacity across plants where cells are coming off lines, and aggregate announced targets north of 140 GWh by year-end. The gap between nameplate and contractable supply is not a data-quality problem. Nearly all accepted US cell output routes through captive channels: Panasonic to Tesla, SK to Ford, LGES Holland through Vertech. A buyer seeking direct cell supply outside an integrated product or anchor-customer relationship faces a pool closer to zero than to 20.
Separately, from public filings, no US cell plant has simultaneously documented accepted output, facility-level 45X credit claims, and PFE compliance. The verified-and-compliant figure is 0 GWh. Credit claims appear only in corporate-level earnings disclosures, without plant-level allocation.
The facility-level evidence behind these figures follows in the Contractable Supply Inventory.
~100 GWh nameplate across US plants in some stage of production: Panasonic NV+KS, LGES Holland, SK Commerce, Fixx Smyrna, StarPlus Kokomo, plus recent first-production starts at Spring Hill and L-H Ohio
~20 GWh with evidence of accepted recurring shipments: LGES Holland (~17 GWh ESS LFP) and Fixx/AESC Smyrna (~3 GWh est., ESS LFP)
Near 0 GWh freely contractable without displacing an anchor customer or buying through an integrated channel
0 GWh with publicly verified 45X + PFE compliance at facility level
Ramp reality: IEA's 2026 Global EV Outlook finds most cell plants take 5+ years from SOP to approach nameplate output; the gap between announced and contractable capacity reflects this timeline, not just captive allocation