BantuNomics· Verbs Enter the ecosystem
The Bantu Verb domain · ABS-300 · BTS-V100

The Bantu verb, as a generative equation.

Not a dataset — the program. Every Bemba verb is built from a 9-slot machine. We share the equation, the full morpheme inventory, an interactive builder, a progressive tutorial, and consented 6-in-1 code-switch recordings — a living platform that grows as native speakers record.

S1[PreInit] + S2[SM] + S3[TAM] + S4[OM] + S5[Root] + S6[Ext1] + S7[Ext2] + S8[FV] + S9[Loc]
S1 PreInit S2 SM S3 TAM S4 OM S5 Root S6 Ext1 S7 Ext2 S8 FV S9 Loc
53Verb types
49Phonological rules
3630Recorded takes · live
6Languages recording
The program
Four ways to use it
Standard

The Equation & Builder

The 9-slot verb equation, the full S1–S9 morpheme inventory, and an interactive builder — pick morphemes, watch the surface form, gloss and translation assemble live.

Open the builder →
The program

The Verb Tutorial

A progressive course that teaches the whole Bemba verb system — build and decompose every verb type, with phonology notes and exercises. Preview is open; the full course is licensed.

Start the tutorial →
Living data

6-in-1 Recordings

One consented Bantu→English code-switch take = six datasets. Hear it proven on a real recording, with live coverage from the field.

See the six datasets →
Structure

Language VT Structure

Pick a language and see its verb table of contents — every VT code, its slot signature, and how much is curated vs recorded (as-of, live).

Browse the structure →
Access
How a lab licenses it

Try the whole engine for free. When you're ready to license, one subscription spans every BantuNomics domain and language — not just verbs.

Public
$0 · no login
  • The 9-slot equation + interactive builder
  • The live engine — generate & decompose (API + MCP)
  • Tutorial preview + dataset-shape teaser
Prove the gap for free. Your agents can discover and call it with no key.
Evaluation
$0 · invite
  • The full 37-chapter interactive tutorial
  • Engine at 50× batch, full API + MCP
  • Sampled consented audio (Bemba)
Verify on a real slice with one token — across every domain, not just verbs.
Pilot
Talk to us · fixed term
  • An agreed language set, bigger caps (25 clips/lang)
  • Measured on your own held-out data
  • The curation moat
License an agreed set for a fixed term, scored on your evaluation.
Whole program
Full Annual Subscription
Talk to us
  • Every domain, every language — syllables, verbs, tone, health, numbers, and each one added next
  • The full consented audio corpus + all bulk exports & APIs
  • Uncapped engine — everything curated while you subscribe
Not a dataset — a subscription to a platform that grows.
One recording, six datasets

We don't sell verb data. We share the machine that makes it.

The generative engine specifies every verb; the recordings prove it in the mouths of native speakers. License the program across the Bantu family.

The catalogue → Get evaluation access
Access

Start where you are.

Prove it free, validate it on your own data, or license the whole program.

01 · Free
Evaluation

Score your model against the foundational layer — the Alphabet Test and the L26 Lite suite, with saved results. Self-serve, no cost.

Start evaluating →
Most labs start here
02 · 75 days
Validation Pilot

Three languages you choose — and everything we hold for them. Measured on your own held-out data.

Scope a pilot →
03 · Program
Full Annual Subscription

Every product, every language, the full consented corpus — and everything curated while you're subscribed.

Talk to us →