Enterprise multi-model usage share (Menlo)
Share of enterprises using more than one foundation model, from Menlo Ventures' enterprise surveys (60% in the 2023 wave, restated in 2024); the 2024 wave states 'three or more models' without a share and the 2025 mid-year update gives the share that switched vendors (11%), shown alongside. a16z's annual CIO survey of the Global 2000, a second survey house, gives the share using three or more model families, also shown alongside.
Reason through thisThe reading
Sixty percent multi-homed in 2023 and the pattern has only spread since, but Menlo stopped publishing the share; the 11% vendor-switch figure says buyers add models rather than replace them.
1 reading drawn from menlo.enterprise.multi_model_share.pt, one of 4 series listed under How it is measured; 5 observations in the evidence.
Source: Menlo Ventures, State of Generative AI in the Enterprise. Menlo Ventures, '2025: The State of Generative AI in the Enterprise'.
The numbers
| as of | reading | value | record |
|---|---|---|---|
| 2023-11-10 | enterprise | 60.0% | reported |
Why it matters
Multi-homing is the model layer's pricing power in one number: buyers who route across providers keep the surplus; buyers locked to one hand it over.
What cuts against it
Menlo's numeric share is a 2023 figure and its later waves changed the question; a16z's CIO survey, shown alongside, asks a stricter question (three or more model families, testing included) of the Global 2000 only.
How it is measured
- series
menlo.enterprise.multi_model_share.pt - series
menlo.enterprise.models_deployed_typical.pt - series
menlo.enterprise.vendor_switch_share.pt - series
a16z.enterprise.three_plus_model_families_share.pt - source Andreessen Horowitz (a16z) essays · default tier 6 · a16z; short quotation
- source Menlo Ventures, State of Generative AI in the Enterprise · default tier 6 · Report
- proxy behaviour, market_structure · cadence annual
Annual waves; a move under five points is survey noise. More multi-homing disperses the model layer's pricing power.
Applied to menlo.enterprise.multi_model_share.pt.
Latest readings
- 2023-11-10 enterprise · 60.0%as of 2023-11-10
Why this status, and its history
It has 1 reading and its direction rule compares 3, so it reads emerging rather than not yet measurable: the site uses not yet measurable only for what cannot be measured at all. The status turns into a direction once enough readings exist.
- 2026-09-27not yet measurable to emergingconf 45 → 45 · Claude (model)
It has 1 reading and its direction rule compares 3, so it reads emerging rather than not yet measurable: the site uses not yet measurable only for what cannot be measured at all. The status turns into a direction once enough readings exist.
Weighed against it: None needed: no reading has changed; this corrects the word used for too few readings.
source1 - 2026-09-10unmeasured to not yet measurableconf — → 45 · the evaluator (automatic)
First reading. Menlo's 2023 enterprise survey put multi-model use at 60% (restated January 2024); later waves state the pattern without a share and the 2025 mid-year update gives an 11% vendor-switch rate instead. One numeric point, so not yet measurable.
source1
Why this confidence
Rests on 5 observations from Andreessen Horowitz (a16z) essays, Menlo Ventures, State of Generative AI in the Enterprise; best evidence tier 6, grade B; current.
Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; 0–49 limited or vague evidence.
Predictions that lean on this
- Too early to tell
- Too early to tell