BEActive

Will Bloom Energy announce a new named hyperscaler customer (beyond Oracle) by October 31, 2026?

Resolves October 31, 2026(185d)

IG: 0.79

Current Prediction

45%

Likely No

Model Agreement92%

Predictions9 runs

Last UpdatedApril 29, 2026

Prediction Distribution

0%25%50%75%100%

opus

sonnet

haiku

Range: 42% – 55%Aggregate: 45%

Individual Predictions(9 runs)

opusRun 1

42%

The bar is strict — requires named announcement, not vague 'leading hyperscaler' language. Hyperscalers historically prefer not to be named in equipment supplier press releases. However, the AI power crisis narrative is strong, and Oracle's success demonstrates the technology works at hyperscaler scale. Two earnings cycles fall in the window. The 800V DC advantage is increasingly recognized. Probability moderate but the NAMED requirement is a meaningful constraint.

NAMED requirement is strictHyperscalers prefer anonymity in supplier PRAI power crisis narrative supports announcementsTwo earnings cycles in windowOracle proof point exists

opusRun 2

50%

Tail-risk frame: hyperscaler procurement timelines are 12-24 months. Active Bloom RFPs in late 2024/early 2025 could close to public announcement in mid-to-late 2026. However, hyperscalers may route through Equinix or Brookfield rather than direct, which would resolve NO. Also possible: a single hyperscaler announces and then second-mover dynamics drive others to make their own announcements. Roughly coin flip.

Procurement timelines align with windowRouting through co-location possibleSecond-mover dynamics could cluster announcementsGenuine uncertainty

opusRun 3

45%

Bear-leaning: capacity constraints (2GW ramp through end of 2026) may mean Bloom is pacing wins rather than maximizing public announcements. The seven-channel framework explicitly includes Brookfield and co-location — these may be the primary 2026 channels with hyperscalers entering as named customers more aggressively in 2027. Modestly below 50%.

Capacity constraints favor pacingSeven-channel mix may not include direct hyperscaler in 2026Hyperscalers may route through co-location2027 more likely than 2026

sonnetRun 1

50%

Balanced view: AI capex environment is maximally favorable for fuel cell power adoption, and Bloom's 800V DC advantage is increasingly recognized. However, the NAMED requirement is strict — many announcements are likely to use anonymous language. Six months and two earnings cycles is enough time for a named announcement if one is in progress. Coin flip seems right.

AI capex maximally favorable800V DC advantage realNAMED requirement strictSix months and two earnings cyclesCoin flip

sonnetRun 2

55%

Slightly bullish: the broader AI infrastructure narrative has matured to where named hyperscaler announcements are increasingly common (e.g., nuclear deals with Microsoft, Amazon's Talen acquisition). The cultural barrier to naming hyperscaler suppliers is lower than 12-18 months ago. Bloom's Q1 commercial momentum and expanded capacity make it plausible at least one of AAPL/AMZN/GOOG/META/MSFT announces a deal in the next 6 months.

AI announcement culture has shiftedNuclear deals demonstrate naming precedentBloom commercial momentum strong6-month window plausible

sonnetRun 3

45%

The base rate matters: in the prior 12 months pre-Q1 2026, zero new named hyperscalers despite AI capex buildup. The next 6 months may continue that pattern. However, the Q1 print was a commercial inflection that may catalyze announcements. Modest probability below 50%.

Prior 12-month base rate zeroQ1 print may catalyze6-month window more demandingModestly below 50%

haikuRun 1

45%

Two salient drivers: AI power demand pulling toward announcements, hyperscaler historical reluctance to be named pushing against. Roughly equal forces, slight lean below 50%.

AI demand pulls towardHyperscaler reluctance pushes againstRoughly balanced

haikuRun 2

50%

Genuine coin flip given the NAMED requirement constraint and the 6-month window.

NAMED requirement strict6-month windowCoin flip

haikuRun 3

42%

Pattern: prior 12 months had zero new named hyperscalers despite AI buildup. Suggests structural reluctance. Slightly below 50%.

Zero base rate prior 12 monthsStructural reluctance patternBelow 50%

Resolution Criteria

Resolves YES if Bloom Energy publicly announces, via press release, 8-K, or earnings call disclosure, a new named hyperscaler customer (Amazon, Google, Microsoft, Meta, Apple, ByteDance, or comparably-scaled hyperscale cloud operator) with a fuel cell deployment, PPA, or supply agreement by October 31, 2026. The announcement must name the customer specifically — references to 'a leading hyperscaler' or 'a major cloud provider' without naming do NOT count. Resolves NO otherwise.

Resolution Source

Bloom Energy press release, SEC filing, or earnings call transcript

Source Trigger

800V DC adoption timeline is uncertain — competitive position depends on hyperscaler PPA capacity wins. Tracks whether the 800V DC native advantage and AI demand thesis convert to additional named hyperscaler customer wins beyond Oracle.

moat-mapperCOMPETITIVE_POSITIONHIGH

View BE Analysis

Full multi-lens equity analysis