Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Hume AI / EVI

EVI 3

English voice conversations with a native speech-language model or a supplemental language model.

deprecatedClosed weightsVoice conversation
Official documentation (opens in a new tab)Reviewed 13 Oct 2026
7 of 18 specification fields collected
Context windowNot collectedAwaiting source review
Maximum outputNot collectedAwaiting source review
Input billingNot collectedHume API
Output billingNot collectedNo inferred price

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
Maximum output
Not collected
Reasoning
Not collected
Tool calling
Not collected

Tool use depends on a supported supplemental or custom language model. It is not established as a general native speech-model capability.

Structured output
Not collected
Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Architecture
Native speech-language model; optional supplemental LLM.Source for Architecture (opens in a new tab)
Context caching
Not collected
Usage metering
Subscription includes TTS, EVI and voice features, with plan allowances and overages. Current numeric rates remain unverified because the public pricing page redirects to the homepage.Source for Usage metering (opens in a new tab)
Voice controls
Explicit voice selection required; uses Voice Library or custom voices, not legacy EVI 1/2 voice IDs.Source for Voice controls (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
3

evi_version in the selected Config; omission of config_id uses EVI 3.

Access
Hume API key required; subscription and account limits apply.
Regions
Deployment region and customer-specific residency are not established by the reviewed endpoint documentation.
Request mode
Bidirectional WebSocket conversation; configuration determines the voice and supplemental model.Source for Request mode (opens in a new tab)
Tool calling
SupportedSource for Tool calling (opens in a new tab)

Requires a supported supplemental or custom LLM; do not interpret this as native hume-evi-3 tool support.

Native model options
hume-evi-3, hume-evi-3-websearchSource for Native model options (opens in a new tab)

Named native options grouped under EVI 3. Web search is a service option; independent checkpoint weights are not established.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Privacy documentation is internally inconsistent: the EVI-specific section says retention and possible training are enabled by default, while the general API FAQ says API data is not used for training. Verify the account settings and contract; no blanket no-training guarantee is asserted.

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.

Hume API

3
Not verified

No verified provider-side caching policy or cache rate. Continuation, stored chats and downloaded audio do not establish prompt caching.

How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Not collected
Limits & data boundaries
Not collected

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

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

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

Inferred for this exact named Hume API release/service from provider documentation and its public weight inventory. No weights for this identity were identified; public TADA releases are excluded. A service record does not assert a separately disclosed checkpoint or parameter count.

Classification evidence (opens in a new tab)

Lifecycle dates

API availability announced18 Jul 2025Source (opens in a new tab)
TTS and EVI sunset announced2 Oct 2026Source (opens in a new tab)
Scheduled API access end at 12:01 a.m. EST13 Nov 2026Source (opens in a new tab)

Sources reviewed