Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Namu / Namu speech families

Namu Aïr

Announced Hausa conversation workflow for phone calls and in-app assistance, with interruption handling and human handover.

announcedClosed-source managed componentVoice conversation
Official documentation (opens in a new tab)Reviewed 16 Oct 2026
8 of 22 specification fields collected
Context windowNot collectedAwaiting source review
Maximum outputNot collectedAwaiting source review
Input billingSee ratesNamu announced service
Output billingSee ratesPublished units and conditions below

Specifications

Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.

Context / input limit
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Maximum input tokens
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Maximum output
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Parameters
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Release date
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Knowledge cutoff
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Tool calling
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Structured output
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Streaming
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Batch processing
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Fine-tuning
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Reasoning
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Context caching
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Base model
Not collected

No reviewed model-specific contract. Unknown is not unsupported.

Availability
Announced family; operational availability unverified. Public playground source explicitly says its backend endpoints do not exist yet, and the inspected repository tree has no playground handlers. The live deployment may differ. Terms describe active development and demonstration/evaluation use; no inference was attempted.Source for Availability (opens in a new tab)
Language coverage
Hausa across eight dialects; French advertised on request.Source for Language coverage (opens in a new tab)
Deployment options
Phone call or in app; advertised to work over ordinary voice calls down to 2G.Source for Deployment options (opens in a new tab)
Interactive controls
Caller interruption and correction; handover to a person at any point.Source for Interactive controls (opens in a new tab)
Usage metering
Advertised $0.05 per minute; telephony inclusion, billing increments and production availability unverified.Source for Usage metering (opens in a new tab)

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

API model ID
Not collected

Family display name and UI route are not immutable native checkpoint IDs.

Access
Contact-sales offering. Public source proposes /api/playground/agent but explicitly says handlers do not yet exist. This is not a verified callable API. Live deployment parity unverified.
Regions
No production inference region established. Public-service privacy mentions transfers to Niger, United States and provider countries.

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 16 Oct 2026
CategoryUsageRateUnit
Advertised price; availability unverifiedAdvertised service rate1$0.05minute

Rate conditions

  1. Model-page advertised rate. Public source describes unimplemented playground handlers; actual service availability, billing rounding, minimum charges, taxes and plan conditions unverified. Not a token price or a verified production quote.
Pricing source (opens in a new tab)

Data & enterprise terms

Reviewed privacy policy covers the public website and Playground and explicitly excludes unreleased products. Training/evaluation use is permitted with permission or where otherwise allowed by law, with withdrawal and deletion requests. Retention is purpose-based without an exact TTL. Transfers may involve Niger, the United States and other provider countries. This is not a blanket no-training, zero-retention or production API residency guarantee.

Read the service policy (opens in a new tab)

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

Namu announced service

Not verified

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

How it works
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

Minimum reusable input
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

Lifetime & refresh
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

How to enable it
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

Usage fields
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

Billing rules
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

Limits & data boundaries
Not collected

No model-specific inference caching contract established. Conversation history and website language-preference storage are not prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Caching reviewed 16 Oct 2026

What the measurements say

Conversational reply gap

Namu · Provider-reported; not independently verified
0.6seconds

Caller stops to agent starts; percentile, call count, channel and hardware not provided. No complete evaluation artifact, versioned serving checkpoint or independently reproduced result reviewed. Operational availability remains unverified.

Evaluation source (opens in a new tab)

Calls finished without a person

Namu · Provider-reported; not independently verified
91%

Provider describes completed calls without human assistance; task mix, denominator and success adjudication not provided. No complete evaluation artifact, versioned serving checkpoint or independently reproduced result reviewed. Operational availability remains unverified.

Evaluation source (opens in a new tab)

Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.

The record behind the model

Why it is included

Named proprietary service family advertised under Namu service terms, which reserve technology rights and prohibit extracting weights. A private native checkpoint and its ancestry have not been established. Kept as a managed family, not a verified closed-weight foundation-model release; public Tiny Aya and educational MMS derivatives are separate projects.

Classification evidence (opens in a new tab)

Sources reviewed