---
title: "List Transcription Models"
method: GET
path: "/v3/audio/transcriptions/models"
tags: ["Audio"]
---

# List Transcription Models

`GET /v3/audio/transcriptions/models`

## Request

- Base URL: `https://api.edenai.run`
- URL: `https://api.edenai.run/v3/audio/transcriptions/models`
- Auth: none declared

List speech-to-text models available in the caller's region.

## Query parameters

- `view` 'endpoints' | 'models': How to group the listing. 'endpoints' (default) is one entry per provider/model id, unchanged. 'models' is one entry per routable model name with its provider endpoints nested underneath, plus one per routing policy you can call — send that name as `model` to let Eden AI choose the provider.
- `zero_data_retention` boolean, nullable: true keeps only models whose provider keeps no prompts or outputs after the response (Zero Data Retention); false keeps only those known to keep them. Models not verified yet (`zero_data_retention: null`) match neither.

## Response `200`

Successful Response

- union
  - ListModelsResponse: List models response.
    - `object` 'list': Object type
    - `data` ModelObject[], required: List of models
      - `id` string, required: Model identifier (e.g., 'openai/gpt-4')
      - `object` 'model': Object type
      - `created` integer, required: Unix timestamp of model creation/release
      - `owned_by` string, required: Provider/organization that owns the model
      - `model_name` string, required: Model name without provider prefix
      - `context_length` integer, nullable: Maximum context length in tokens
      - `description` string, nullable: Model description
      - `source` string, nullable: Model source URL
      - `capabilities` Capabilities: Model capability flags. Unknown flags are preserved so consumers keep working when new capabilities are introduced without a coordinated release.
        - `input_modalities` string[]
        - `output_modalities` string[]
        - `supports_reasoning` boolean
        - `supports_web_search` boolean
        - `supports_batch` boolean: true when this entry passes the model checks of `/v3/batches`; false when a batch on it is refused (a provider batches do not run on, an id with a region tag or `:` suffix, a model served through region variants, its default variant included, no published batch rate, cached input priced with no cached batch rate, or long-context bands, audio, or cache writes where the provider's batch rates do not price them). Computed per entry, so the same model can be true on one entry and false on another. On a `view=models` entry it is true when any endpoint is; batch lines take `<provider>/<model>` ids, so name an `endpoints[].id` that reads true. It does not read the caller or the request: a create on a true entry can still be refused for the caller's own key for the provider (400), a routing policy on the key (400), an empty balance or spent member budget (402), zero data retention (403), the provider's region or provider gates (451), repeated content moderation refusals (429) or while batches are paused (503). The listing is cached, so a catalog change can reach the create before it reaches this flag.
        - `supports_tool_choice` boolean
        - `supports_computer_use` boolean
        - `supports_prompt_caching` boolean
        - `supports_response_schema` boolean
        - `supports_system_messages` boolean
        - `supports_function_calling` boolean
        - `supports_native_streaming` boolean
        - `supports_assistant_prefill` boolean
        - `supports_embedding_image_input` boolean
        - `supports_parallel_function_calling` boolean
      - `pricing` object: Pricing information after any applicable discounts have been applied
      - `list_pricing` object: Provider list pricing, before any discounts are applied
      - `discount` number, nullable: Discount applied to the model (0-1 range)
      - `regions` RegionObject[]: Regions where this model is available
        - `code` string, required: Region code (e.g., 'us-east-1')
        - `name` string, required: Region display name
      - `zero_data_retention` boolean, nullable: true when calls to this entry keep no prompts or outputs after the response (Zero Data Retention), under Eden AI's terms with its provider: the model's own terms when they differ, else the provider's; a customer-hosted model is always true. false when they are kept, null when not verified yet.
      - `alias_of` string, nullable: Set when this entry is an alias (e.g. 'gemini-pro-latest'): the real model_id it currently resolves to. None for concrete models.
  - ListModelsWithEndpointsResponse: `?view=models` response: routable names, each with its provider endpoints. Named for the public vocabulary rather than the internal one: Pydantic class names become OpenAPI schema titles and generated SDK classes, and a caller never needs the word "canonical" -- to them these are simply models, and the rows behind them endpoints.
    - `object` 'list': Object type
    - `data` ModelWithEndpoints[], required: List of routable models
      - `id` string, required: The routable model name, with no provider prefix (e.g. 'gpt-oss-120b'), or a routing policy you can call (`policy/{name}`). Send this as `model` to let Eden AI choose which provider serves it.
      - `object` 'model': Object type
      - `created` integer: Unix timestamp of the newest endpoint serving this model
      - `owned_by` string: Who authored the model, independent of who sells it
      - `mode` string, nullable: What the model does: 'chat', 'embedding', 'stt', 'tts' or 'image_generation'. Lets a caller tell an LLM from a voice.
      - `context_length` integer, nullable: The largest context window among the endpoints. A request is only routed to endpoints whose window fits it.
      - `capabilities` Capabilities: Model capability flags. Unknown flags are preserved so consumers keep working when new capabilities are introduced without a coordinated release.
        - `input_modalities` string[]
        - `output_modalities` string[]
        - `supports_reasoning` boolean
        - `supports_web_search` boolean
        - `supports_batch` boolean: true when this entry passes the model checks of `/v3/batches`; false when a batch on it is refused (a provider batches do not run on, an id with a region tag or `:` suffix, a model served through region variants, its default variant included, no published batch rate, cached input priced with no cached batch rate, or long-context bands, audio, or cache writes where the provider's batch rates do not price them). Computed per entry, so the same model can be true on one entry and false on another. On a `view=models` entry it is true when any endpoint is; batch lines take `<provider>/<model>` ids, so name an `endpoints[].id` that reads true. It does not read the caller or the request: a create on a true entry can still be refused for the caller's own key for the provider (400), a routing policy on the key (400), an empty balance or spent member budget (402), zero data retention (403), the provider's region or provider gates (451), repeated content moderation refusals (429) or while batches are paused (503). The listing is cached, so a catalog change can reach the create before it reaches this flag.
        - `supports_tool_choice` boolean
        - `supports_computer_use` boolean
        - `supports_prompt_caching` boolean
        - `supports_response_schema` boolean
        - `supports_system_messages` boolean
        - `supports_function_calling` boolean
        - `supports_native_streaming` boolean
        - `supports_assistant_prefill` boolean
        - `supports_embedding_image_input` boolean
        - `supports_parallel_function_calling` boolean
      - `pricing` object, nullable: The full price list of the cheapest endpoint a call without a provider can reach, named by `pricing_endpoint`. A call routed elsewhere, or to a fallback, is billed at that endpoint's prices. null when no endpoint is priced.
      - `pricing_endpoint` string, nullable: The endpoint whose prices `pricing` shows.
      - `regions` RegionObject[]: Every region at least one endpoint runs in.
        - `code` string, required: Region code (e.g., 'us-east-1')
        - `name` string, required: Region display name
      - `endpoints` ModelObject[], required: Every provider endpoint serving this model, each with its own pricing, context length, capabilities and regions. Empty for a routing policy: which of its targets serves a call is decided per request.
        - `id` string, required: Model identifier (e.g., 'openai/gpt-4')
        - `object` 'model': Object type
        - `created` integer, required: Unix timestamp of model creation/release
        - `owned_by` string, required: Provider/organization that owns the model
        - `model_name` string, required: Model name without provider prefix
        - `context_length` integer, nullable: Maximum context length in tokens
        - `description` string, nullable: Model description
        - `source` string, nullable: Model source URL
        - `capabilities` Capabilities: Model capability flags. Unknown flags are preserved so consumers keep working when new capabilities are introduced without a coordinated release.
          - `input_modalities` string[]
          - `output_modalities` string[]
          - `supports_reasoning` boolean
          - `supports_web_search` boolean
          - `supports_batch` boolean: true when this entry passes the model checks of `/v3/batches`; false when a batch on it is refused (a provider batches do not run on, an id with a region tag or `:` suffix, a model served through region variants, its default variant included, no published batch rate, cached input priced with no cached batch rate, or long-context bands, audio, or cache writes where the provider's batch rates do not price them). Computed per entry, so the same model can be true on one entry and false on another. On a `view=models` entry it is true when any endpoint is; batch lines take `<provider>/<model>` ids, so name an `endpoints[].id` that reads true. It does not read the caller or the request: a create on a true entry can still be refused for the caller's own key for the provider (400), a routing policy on the key (400), an empty balance or spent member budget (402), zero data retention (403), the provider's region or provider gates (451), repeated content moderation refusals (429) or while batches are paused (503). The listing is cached, so a catalog change can reach the create before it reaches this flag.
          - `supports_tool_choice` boolean
          - `supports_computer_use` boolean
          - `supports_prompt_caching` boolean
          - `supports_response_schema` boolean
          - `supports_system_messages` boolean
          - `supports_function_calling` boolean
          - `supports_native_streaming` boolean
          - `supports_assistant_prefill` boolean
          - `supports_embedding_image_input` boolean
          - `supports_parallel_function_calling` boolean
        - `pricing` object: Pricing information after any applicable discounts have been applied
        - `list_pricing` object: Provider list pricing, before any discounts are applied
        - `discount` number, nullable: Discount applied to the model (0-1 range)
        - `regions` RegionObject[]: Regions where this model is available
          - `code` string, required: Region code (e.g., 'us-east-1')
          - `name` string, required: Region display name
        - `zero_data_retention` boolean, nullable: true when calls to this entry keep no prompts or outputs after the response (Zero Data Retention), under Eden AI's terms with its provider: the model's own terms when they differ, else the provider's; a customer-hosted model is always true. false when they are kept, null when not verified yet.
        - `alias_of` string, nullable: Set when this entry is an alias (e.g. 'gemini-pro-latest'): the real model_id it currently resolves to. None for concrete models.
      - `endpoint_count` integer, required: How many provider endpoints serve this model

## Other responses

- `422`: Validation Error

---

[API](https://skmtc.dev/edenai/apis/eden-ai-api-v3.md) - [All operations](https://skmtc.dev/edenai/apis/eden-ai-api-v3/llms.txt) - [OpenAPI document](https://skmtc.dev/edenai/apis/eden-ai-api-v3/revisions/20ddfd49a3f4?raw)