> ## Documentation Index > Fetch the complete documentation index at: https://veniceai-mintlify-6ce01df5.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # List Models > Returns a list of available models supported by the Venice.ai API across text, image, audio, video, and related inference types. ## Quality-Tier Pricing For image models that accept the optional `quality` parameter (currently `gpt-image-2` and `gpt-image-2-edit`), the response exposes a per-quality price matrix under `model_spec.pricing.quality`. Each top-level key is a resolution tier (`1K`, `2K`, `4K`) and each nested key is a quality level (`low`, `medium`, `high`) carrying its own `usd` and `diem` price: ```json theme={"system"} "pricing": { "resolutions": { "1K": { "usd": 0.27, "diem": 0.27 }, "2K": { "usd": 0.51, "diem": 0.51 }, "4K": { "usd": 0.84, "diem": 0.84 } }, "quality": { "1K": { "low": { "usd": 0.02, "diem": 0.02 }, "medium": { "usd": 0.07, "diem": 0.07 }, "high": { "usd": 0.26, "diem": 0.26 } }, "2K": { "low": { "usd": 0.03, "diem": 0.03 }, "medium": { "usd": 0.13, "diem": 0.13 }, "high": { "usd": 0.50, "diem": 0.50 } }, "4K": { "low": { "usd": 0.05, "diem": 0.05 }, "medium": { "usd": 0.21, "diem": 0.21 }, "high": { "usd": 0.83, "diem": 0.83 } } } } ``` `pricing.resolutions` is the legacy per-image schedule kept for backward compatibility. `pricing.quality` is the per-(resolution, quality) matrix that applies whenever the `quality` parameter is supported. Both fields are kept in the response so clients can detect quality support and surface the matrix in their own UIs. ## Postman Collection For additional examples, please see this [Postman Collection](https://www.postman.com/veniceai/workspace/venice-ai-workspace/folder/38652128-59dfa959-7038-4cd8-b8ba-80cf09f2f026?action=share\&source=copy-link\&creator=38652128\&ctx=documentation). *** ## OpenAPI ````yaml GET /models openapi: 3.0.0 info: description: The Venice.ai API. termsOfService: https://venice.ai/legal/tos title: Venice.ai API version: '20260617.063120' x-guidance: >- Venice.ai is an OpenAI-compatible inference API supporting text, image, audio, and video generation. **Authentication options:** - API Key: Use Bearer token in Authorization header - x402 Wallet: Use USDC credits via EVM or Solana wallet (no account required) **For x402 wallet access:** 1. POST /x402/top-up without headers to get payment requirements 2. Choose one of the returned Base or Solana payment options and sign a USDC payment using the x402 SDK 3. POST /x402/top-up with PAYMENT-SIGNATURE header to add credits 4. Call any inference endpoint with SIGN-IN-WITH-X header **Pricing:** Prepaid credits consumed per request. Check /models for available models and their capabilities. servers: - url: https://api.venice.ai/api/v1 security: - BearerAuth: [] tags: - description: >- Generate speech/audio, transcribe audio, and manage asynchronous audio generation jobs. name: Audio - description: >- Given a list of messages comprising a conversation, the model will return a response. Supports multimodal inputs including text, images, audio (input_audio), and video (video_url) for compatible models. name: Chat - description: List and describe the various models available in the API. name: Models - description: Generate and manipulate images using AI models. name: Image - description: Generate videos using AI models. name: Video - description: List and retrieve character information for use in completions. name: Characters - description: >- Billing and usage analytics. **Beta**: This API is currently in beta and may be unstable. Endpoints, request/response schemas, and behavior may change without notice. name: Billing - description: Proxy JSON-RPC requests to blockchain nodes. Billed per credit. name: Crypto RPC - description: >- Wallet-based API access using the x402 protocol. No API key required — authenticate with an EVM or Solana wallet. **How it works:** 1. **Authenticate** — Send a `SIGN-IN-WITH-X` header (base64-encoded signed SIWX payload) with any request. EVM wallets sign an EIP-4361 SIWE message; Solana wallets sign the Solana SIWX message with Ed25519. See the `siwx` security scheme for the exact format. 2. **Top up** — `POST /x402/top-up` without a payment header returns an `accepts` array with Base and Solana USDC payment options. Choose one entry, sign it using the x402 SDK (`npm install x402`), and re-submit with the `PAYMENT-SIGNATURE` header (the legacy `X-402-Payment` and `X-PAYMENT` names are also accepted). 3. **Use any endpoint** — All inference endpoints (chat, image, audio, video, embeddings) accept `siwx` as an alternative to `BearerAuth`. Charges are deducted from your USDC credit balance. 4. **Monitor balance** — `GET /x402/balance/{walletAddress}` returns your current balance. The `X-Balance-Remaining` response header on inference calls also reports it. **Quick start (5 lines):** ``` import { VeniceClient } from '@venice-ai/x402-client' const venice = new VeniceClient(process.env.WALLET_KEY) await venice.topUp(10) // $10 USDC on a supported x402 rail const res = await venice.chat({ model: 'zai-org-glm-5-1', messages: [{ role: 'user', content: 'Hello!' }] }) ``` **Payment:** USDC on Base (chain ID 8453) or Solana mainnet. Minimum top-up: $5. Alternatively, stake DIEM tokens for daily credits (1 DIEM = $1/day). name: x402 externalDocs: description: Venice.ai API documentation url: https://docs.venice.ai paths: /models: get: tags: - Models summary: /api/v1/models description: >- Returns a list of available models supported by the Venice.ai API across text, image, audio, video, and related inference types. operationId: listModels parameters: - schema: anyOf: - type: string enum: - asr - embedding - image - music - text - tts - upscale - inpaint - video - type: string enum: - all - code description: Filter models by type. Use "all" to get all model types. example: text required: false name: type in: query responses: '200': description: OK content: application/json: schema: type: object properties: data: type: array items: $ref: '#/components/schemas/ModelResponse' description: List of available models object: type: string enum: - list type: anyOf: - type: string enum: - asr - embedding - image - music - text - tts - upscale - inpaint - video - type: string enum: - all - code description: Type of models returned. example: text required: - data - object - type '500': description: An unknown error occurred content: application/json: schema: $ref: '#/components/schemas/StandardError' security: - {} - BearerAuth: [] components: schemas: ModelResponse: type: object properties: context_length: type: number description: >- The context length (maximum input tokens) supported by the model. Only present for text models. This is a standard OpenAI-compatible field that mirrors model_spec.availableContextTokens for client compatibility. example: 131072 created: type: number description: Release date on Venice API example: 1699000000 id: type: string description: Model ID example: zai-org-glm-5-2 model_spec: type: object properties: availableContextTokens: type: number description: >- The context length supported by the model. Only applicable for text models. example: 1000000 maxCompletionTokens: type: number description: >- The maximum number of completion tokens the model can generate. Use this to know the upper bound for the max_completion_tokens request parameter. Only applicable for text models. example: 24000 beta: type: boolean description: >- Is this model restricted to beta users only? If true, only users with beta access can use this model example: false betaModel: type: boolean description: Is this model in beta status? example: false privacy: type: string enum: - private - anonymized description: >- The privacy mode of the model. Private models have zero data retention. Anonymized models Venice can not guarantee privacy on, but requests are not affiliated with a user example: private regionRestrictions: type: array items: type: string description: >- Country codes where this model is intended to be available. Only present for models with region restrictions metadata. example: - US deprecation: type: object properties: autoRemap: type: boolean description: >- When true, Venice may automatically remap API requests for this model ID to replacementModelId instead of returning an error. example: false date: type: string description: >- Legacy ISO 8601 instant aligned with the deprecation sunset used in response headers (`x-venice-model-deprecation-date`). Prefer startsAt / removesAt for new integrations. example: '2025-03-01T00:00:00.000Z' removesAt: type: string description: >- ISO 8601 instant when this model ID is omitted from public GET /models listings (same value as `date` today). Consumers should treat the current wall-clock time as past this instant when deciding whether the model remains listed. example: '2025-04-01T00:00:00.000Z' replacementModelId: type: string description: >- Suggested public API model ID to migrate to, when one exists. example: llama-3-3-70b startsAt: type: string description: >- ISO 8601 instant when deprecation warnings and documentation should be considered active for this model. example: '2025-03-01T00:00:00.000Z' required: - autoRemap - date - removesAt description: >- Deprecation information for the model. Only present for models scheduled to be retired example: autoRemap: false date: '2025-03-01T00:00:00.000Z' removesAt: '2025-03-01T00:00:00.000Z' capabilities: type: object properties: optimizedForCode: type: boolean description: Is the LLM optimized for coding? example: true quantization: type: string enum: - fp4 - fp8 - fp16 - bf16 - int8 - int4 - not-available description: The quantization type of the running model. example: fp8 supportsFunctionCalling: type: boolean description: Does the LLM model support function calling? example: true supportsAudioInput: type: boolean description: Does the LLM support audio input? example: false supportsReasoning: type: boolean description: >- Does the model support reasoning with blocks of output. example: true supportsReasoningEffort: type: boolean description: >- Does the model support the reasoning_effort parameter to control reasoning depth. example: true reasoningEffortOptions: type: array items: type: string enum: - none - minimal - low - medium - high - xhigh - max description: >- Supported reasoning_effort values for this model. Only present when supportsReasoningEffort is true. The "none" option means reasoning can be disabled. example: - none - low - medium - high defaultReasoningEffort: type: string enum: - none - minimal - low - medium - high - xhigh - max description: >- Default reasoning_effort value used when the request does not specify one. Only present when supportsReasoningEffort is true. example: medium supportsResponseSchema: type: boolean description: >- Does the LLM model support response schema? Only models that support function calling can support response_schema. example: true supportsMultipleImages: type: boolean description: >- Does the LLM support multiple images in a single request? Only applicable when supportsVision is true. example: true maxImages: type: number description: >- Maximum number of images supported per request. Only present when supportsMultipleImages is true. example: 10 supportsVision: type: boolean description: Does the LLM support vision? example: true supportsVideoInput: type: boolean description: Does the LLM support video input? example: true supportsWebSearch: type: boolean description: Does the LLM model support web search? example: true supportsLogProbs: type: boolean description: Does the LLM model support logprobs parameter? example: true supportsTeeAttestation: type: boolean description: >- Does the model run inside a Trusted Execution Environment (TEE) with hardware attestation? When true, use GET /tee/attestation and GET /tee/signature to cryptographically verify that inference occurred inside a genuine TEE. example: false supportsE2EE: type: boolean description: >- Does the model support End-to-End Encryption (E2EE)? When true, clients can encrypt prompts using the TEE public key from attestation, and responses are returned encrypted. Requires supportsTeeAttestation to also be true. example: false supportsXSearch: type: boolean description: >- Does the model support xAI's native X Search (web + X/Twitter search)? When true, you can use venice_parameters.enable_x_search to activate real-time search powered by xAI. example: false required: - optimizedForCode - quantization - supportsFunctionCalling - supportsAudioInput - supportsReasoning - supportsReasoningEffort - supportsResponseSchema - supportsMultipleImages - supportsVision - supportsVideoInput - supportsWebSearch - supportsLogProbs - supportsTeeAttestation - supportsE2EE - supportsXSearch additionalProperties: false description: Text model specific capabilities. constraints: anyOf: - type: object properties: aspectRatios: type: array items: type: string description: >- Supported aspect ratio options for this model. Only present for models that support aspect ratio selection. example: - '1:1' - '16:9' - '9:16' - '3:2' - '2:3' defaultAspectRatio: type: string description: >- The default aspect ratio for this model. Only present for models that support aspect ratio selection. example: '1:1' defaultResolution: type: string description: >- The default resolution for this model. Only present for models that support resolution selection. example: 1K resolutions: type: array items: type: string description: >- Supported resolution options for this model. Only present for models that support resolution selection. example: - 1K - 2K - 4K defaultQuality: type: string enum: - low - medium - high description: >- The default quality for this model. Only present for models that accept the `quality` request parameter (currently GPT Image 2). example: high qualities: type: array items: type: string enum: - low - medium - high description: >- Supported quality options for this model. Only present for models that accept the `quality` request parameter (currently GPT Image 2). example: - low - medium - high promptCharacterLimit: type: number description: The maximum supported prompt length. example: 2048 steps: type: object properties: default: type: number description: The default steps value for the model example: 25 max: type: number description: The maximum supported steps value for the model example: 50 required: - default - max widthHeightDivisor: type: number description: >- The requested width and height of the image generation must be divisible by this value. example: 8 required: - promptCharacterLimit - steps - widthHeightDivisor description: Constraints that apply to image models. title: Image Model Constraints - type: object properties: temperature: type: object properties: default: type: number description: The default temperature value for the model example: 0.7 required: - default top_p: type: object properties: default: type: number description: The default top_p value for the model example: 0.9 required: - default frequency_penalty: type: object properties: default: type: number description: The default frequency_penalty value for the model example: 0 required: - default presence_penalty: type: object properties: default: type: number description: The default presence_penalty value for the model example: 0 required: - default repetition_penalty: type: object properties: default: type: number description: The default repetition_penalty value for the model example: 1.05 required: - default required: - temperature - top_p description: Constraints that apply to text models. title: Text Model Constraints - type: object properties: aspect_ratios: type: array items: type: string description: >- The aspect ratios supported by the model. Empty array means the model does not support a defined aspect ratio. example: - '16:9' - '9:16' resolutions: type: array items: type: string description: >- The resolutions supported by the model. Empty array means the model does not support a defined resolution. example: - 1080p - 720p - 480p durations: type: array items: type: string description: >- The durations supported by the model. Empty array means the model does not support a defined duration. example: - 5s - 10s - 15s - 20s - 30s model_type: type: string enum: - image-to-video - text-to-video - video description: The type of video model. example: image-to-video audio: type: boolean description: Does the model support audio generation? example: true audio_configurable: type: boolean description: >- Can audio be enabled or disabled for the video generation? example: true prompt_character_limit: type: number description: >- The maximum supported prompt length for this video model. If not specified, the default is 2500 characters. example: 1000 required: - aspect_ratios - resolutions - durations - model_type - audio - audio_configurable description: Constraints that apply to video models. title: Video Model Constraints - type: object properties: aspectRatios: type: array items: type: string description: >- The aspect ratios supported by this model. Omit the parameter to use the model's default setting. example: - auto - '1:1' - '16:9' - '9:16' promptCharacterLimit: type: number description: The maximum supported prompt length. example: 1500 combineImages: type: boolean description: >- Whether this model supports combining multiple input images. example: true singleImageAspectRatio: type: boolean description: >- If false, output dimensions match the input on single-image edits and `aspect_ratio` is ignored. Multi-image edits are unaffected. Defaults to true. example: true defaultResolution: type: string description: >- The default resolution for this inpaint model. Only present for models that support resolution selection (e.g. GPT Image 2 Edit). example: 1K resolutions: type: array items: type: string description: >- Supported resolution options for this inpaint model. Only present for models that support resolution selection. example: - 1K - 2K - 4K defaultQuality: type: string enum: - low - medium - high description: >- The default quality for this inpaint model. Only present for models that accept the `quality` request parameter (currently GPT Image 2 Edit). example: high qualities: type: array items: type: string enum: - low - medium - high description: >- Supported quality options for this inpaint model. Only present for models that accept the `quality` request parameter (currently GPT Image 2 Edit). example: - low - medium - high required: - aspectRatios - promptCharacterLimit - combineImages description: Constraints that apply to inpaint/edit models. title: Inpaint Model Constraints description: Constraints that apply to this model. description: type: string description: A human-readable description of the model and its capabilities. example: >- Balanced blend of speed and capability. Handles most everyday tasks with reliability. name: type: string description: The name of the model. example: GLM 5.2 modelSource: type: string description: The source of the model, such as a URL to the model repository. example: https://huggingface.co/zai-org/GLM-5.2 offline: type: boolean default: false description: Is this model presently offline? example: false pricing: anyOf: - type: object properties: input: type: object properties: usd: type: number description: USD cost per million input tokens example: 0.7 diem: type: number description: Diem cost per million input tokens example: 7 required: - usd - diem cache_input: type: object properties: usd: type: number description: >- USD cost per million cached input tokens (discounted rate for cache reads) example: 0.35 diem: type: number description: >- Diem cost per million cached input tokens (discounted rate for cache reads) example: 3.5 required: - usd - diem description: >- Optional pricing for cached input tokens (cache reads). Only present for models that support context caching. cache_write: type: object properties: usd: type: number description: >- USD cost per million cache creation tokens (cache writes). For some providers this may be higher than input price. example: 7.5 diem: type: number description: >- Diem cost per million cache creation tokens (cache writes). For some providers this may be higher than input price. example: 75 required: - usd - diem description: >- Optional pricing for cache creation tokens (cache writes). Only present for models where provider charges for cache writes (e.g., Anthropic charges 1.25x input price). output: type: object properties: usd: type: number description: USD cost per million output tokens example: 2.8 diem: type: number description: Diem cost per million output tokens example: 28 required: - usd - diem extended: type: object properties: context_token_threshold: type: number description: >- Input token count above which extended pricing applies example: 200000 input: type: object properties: usd: type: number description: >- USD cost per million input tokens (extended tier) example: 11 diem: type: number description: >- Diem cost per million input tokens (extended tier) example: 11 required: - usd - diem output: type: object properties: usd: type: number description: >- USD cost per million output tokens (extended tier) example: 41.25 diem: type: number description: >- Diem cost per million output tokens (extended tier) example: 41.25 required: - usd - diem cache_input: type: object properties: usd: type: number description: >- USD cost per million cached input tokens (extended tier) example: 1.1 diem: type: number description: >- Diem cost per million cached input tokens (extended tier) example: 1.1 required: - usd - diem cache_write: type: object properties: usd: type: number description: >- USD cost per million cache write tokens (extended tier) example: 13.75 diem: type: number description: >- Diem cost per million cache write tokens (extended tier) example: 13.75 required: - usd - diem required: - context_token_threshold - input - output description: >- Extended pricing for long-context requests exceeding the threshold. When input tokens exceed context_token_threshold, extended rates apply to the entire request. required: - input - output description: >- Token-based pricing for chat models. Models supporting context caching will include cache_input (cache read) pricing. Some providers like Anthropic also charge for cache_write (cache creation) at a premium rate. title: LLM Model Pricing - type: object properties: generation: type: object properties: usd: type: number description: USD cost per image generation (base price) example: 0.01 diem: type: number description: Diem cost per image generation (base price) example: 0.1 required: - usd - diem description: >- Base pricing for image generation. Only present for models without resolution-specific pricing. resolutions: type: object additionalProperties: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem description: >- Resolution-specific pricing. Keys are resolution tiers (e.g. "1K", "2K", "4K"). Only present for models that support resolution selection. When present, replaces `generation`. example: 1K: usd: 0.18 diem: 0.18 2K: usd: 0.24 diem: 0.24 4K: usd: 0.35 diem: 0.35 quality: type: object additionalProperties: type: object properties: low: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem medium: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem high: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem description: >- Per-(resolution, quality) pricing. Only present for models that accept the `quality` request parameter (currently GPT Image 2 / GPT Image 2 Edit). example: 1K: high: usd: 0.27 diem: 0.27 low: usd: 0.02 diem: 0.02 medium: usd: 0.07 diem: 0.07 2K: high: usd: 0.5 diem: 0.5 low: usd: 0.03 diem: 0.03 medium: usd: 0.13 diem: 0.13 4K: high: usd: 0.83 diem: 0.83 low: usd: 0.05 diem: 0.05 medium: usd: 0.21 diem: 0.21 upscale: type: object properties: 2x: type: object properties: usd: type: number description: USD cost for 2x upscale example: 0.02 diem: type: number description: Diem cost for 2x upscale example: 0.2 required: - usd - diem 4x: type: object properties: usd: type: number description: USD cost for 4x upscale example: 0.08 diem: type: number description: Diem cost for 4x upscale example: 0.8 required: - usd - diem required: - 2x - 4x required: - upscale description: Pricing for image generation and upscaling title: Image Model Pricing - type: object properties: input: type: object properties: usd: type: number description: USD cost per million input characters example: 3.5 diem: type: number description: Diem cost per million input characters example: 35 required: - usd - diem required: - input description: Pricing for TTS models title: TTS Model Pricing - type: object properties: per_audio_second: type: object properties: usd: type: number description: USD cost per audio second example: 0.0001 diem: type: number description: Diem cost per audio second example: 0.0001 required: - usd - diem required: - per_audio_second description: Pricing for ASR (speech-to-text) models title: ASR Model Pricing - type: object properties: inpaint: type: object properties: usd: type: number description: USD cost per image edit/inpaint operation example: 0.04 diem: type: number description: Diem cost per image edit/inpaint operation example: 0.04 required: - usd - diem resolutions: type: object additionalProperties: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem description: >- Resolution-specific pricing. Keys are resolution tiers (e.g. "1K", "2K", "4K"). Only present for inpaint models that support resolution selection. example: 1K: usd: 0.36 diem: 0.36 2K: usd: 0.53 diem: 0.53 4K: usd: 0.85 diem: 0.85 quality: type: object additionalProperties: type: object properties: low: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem medium: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem high: type: object properties: usd: type: number description: USD cost for this resolution example: 0.18 diem: type: number description: Diem cost for this resolution example: 0.18 required: - usd - diem description: >- Per-(resolution, quality) pricing. Only present for models that accept the `quality` request parameter (currently GPT Image 2 / GPT Image 2 Edit). example: 1K: high: usd: 0.27 diem: 0.27 low: usd: 0.02 diem: 0.02 medium: usd: 0.07 diem: 0.07 2K: high: usd: 0.5 diem: 0.5 low: usd: 0.03 diem: 0.03 medium: usd: 0.13 diem: 0.13 4K: high: usd: 0.83 diem: 0.83 low: usd: 0.05 diem: 0.05 medium: usd: 0.21 diem: 0.21 required: - inpaint description: Pricing for image editing/inpainting models title: Inpaint Model Pricing - type: object properties: generation: type: object properties: usd: type: number description: USD cost per music generation example: 0.02 diem: type: number description: Diem cost per music generation example: 0.02 required: - usd - diem required: - generation - type: object properties: durations: type: object additionalProperties: type: object properties: usd: type: number description: USD cost for this duration tier example: 0.87 diem: type: number description: Diem cost for this duration tier example: 0.87 min_seconds: type: number description: >- Minimum duration (inclusive) in seconds that falls into this pricing tier example: 1 max_seconds: type: number description: >- Maximum duration (inclusive) in seconds that falls into this pricing tier example: 60 required: - usd - diem - min_seconds - max_seconds required: - durations - type: object properties: per_second: type: object properties: usd: type: number description: USD cost per second of generated music example: 0.005 diem: type: number description: Diem cost per second of generated music example: 0.005 required: - usd - diem required: - per_second - type: object properties: per_thousand_characters: type: object properties: usd: type: number description: USD cost per thousand characters example: 0.01 diem: type: number description: Diem cost per thousand characters example: 0.01 required: - usd - diem required: - per_thousand_characters description: Pricing details for the model traits: type: array items: type: string description: >- Traits that apply to this model. You can specify a trait to auto-select a model vs. specifying the model ID in your request to avoid breakage as Venice updates and iterates on its models. example: - default_code embeddingDimensions: type: number description: >- The native/default number of dimensions in the output embedding vector. Only present for embedding models. example: 1024 maxInputTokens: type: number description: >- Maximum number of input tokens the model accepts per input string. Only present for embedding models. example: 8192 supportsCustomDimensions: type: boolean description: >- Whether the model supports reducing output dimensions via the `dimensions` request parameter. Only present for embedding models that support it. example: true supports_lyrics: type: boolean description: Whether this audio-generation model supports lyrics input. example: true lyrics_required: type: boolean description: >- Whether lyrics input is required for this audio-generation model. example: false supports_force_instrumental: type: boolean description: >- Whether this audio-generation model supports the force_instrumental request parameter. example: true voices: type: array items: type: string description: >- The voices available for this model. Applicable for TTS models and voice-enabled music models. Note: each model has its own set of supported voices. example: - Achernar - Achird - Aiden - Alex - Algenib voice_cloning: type: object properties: mode: type: string enum: - zero_shot - persistent description: >- How the upstream provider implements voice cloning. `zero_shot` re-reads the reference audio on every synthesis call and never derives a persistent voice template (e.g. Chatterbox HD). `persistent` derives a custom voice template upstream that survives across synthesis calls (e.g. MiniMax Speech-02 HD). example: zero_shot accepted_formats: type: array items: type: string description: >- Audio container formats this model accepts as a reference sample on POST /v1/audio/voices. Samples in other containers are rejected with HTTP 400 before any upload. example: - mp3 - wav - flac - mp4 min_sample_seconds: type: number description: >- Recommended minimum length of the reference audio sample, in seconds, for an intelligible clone. example: 5 retention_days: type: number description: >- Days a `vv_` voice handle remains valid against this model. For `persistent` models the upstream provider auto-deletes the cloned voice after this many days without use; each successful TTS request resets the window. For `zero_shot` models this is the storage TTL on the uploaded reference audio; the handle stops working when it expires and the user must re-upload. example: 7 required: - mode - accepted_formats - min_sample_seconds - retention_days additionalProperties: false description: >- Voice-cloning capability. Only present for TTS models whose cloning endpoint is publicly available — pass the model to POST /v1/audio/voices to mint a `vv_` voice handle, then pass that handle back as the `voice` parameter on POST /v1/audio/speech alongside the same `model`. Models with cloning gated behind a private alpha (e.g. MiniMax during the biometric-data legal review) omit this field for non-staff callers; the model itself still appears in the listing. title: TTS Voice Cloning default_voice: type: string description: Default voice for voice-enabled music models. example: Aria supports_language_code: type: boolean description: >- Whether this music model supports an ISO 639-1 language_code parameter. example: true supports_speed: type: boolean description: Whether this music model supports speed adjustment. example: true default_speed: type: number description: Default speed multiplier for this music model. example: 1 min_speed: type: number description: Minimum speed multiplier for this music model. example: 0.25 max_speed: type: number description: Maximum speed multiplier for this music model. example: 4 duration_options: type: array items: type: number description: Available duration options in seconds for this music model. example: - 60 - 120 - 180 - 240 min_duration: type: number description: Minimum duration in seconds for this music model. example: 60 max_duration: type: number description: Maximum duration in seconds for this music model. example: 240 default_duration: type: number description: Default duration in seconds for this music model. example: 60 supported_formats: type: array items: type: string description: Supported audio formats for this music model. example: - mp3 - wav default_format: type: string description: Default audio format for this music model. example: mp3 prompt_character_limit: type: number description: Maximum prompt character limit for this music model. example: 500 min_prompt_length: type: number description: Minimum prompt length for this music model. example: 1 lyrics_character_limit: type: number description: Maximum lyrics character limit for this music model. example: 3000 required: - privacy object: type: string enum: - model description: Object type example: model owned_by: type: string enum: - venice.ai description: Who runs the model example: venice.ai type: type: string enum: - asr - embedding - image - music - text - tts - upscale - inpaint - video description: Model type example: text required: - id - model_spec - object - owned_by - type description: Response schema for model information example: created: 1727966436 id: llama-3.2-3b model_spec: availableContextTokens: 131072 capabilities: optimizedForCode: false quantization: fp16 supportsAudioInput: false supportsFunctionCalling: true supportsLogProbs: true supportsMultipleImages: false supportsReasoning: false supportsReasoningEffort: false supportsResponseSchema: true supportsTeeAttestation: false supportsE2EE: false supportsVision: false supportsVideoInput: false supportsWebSearch: true supportsXSearch: false constraints: temperature: default: 0.8 top_p: default: 0.9 description: >- Compact and efficient model for quick responses and lighter workloads. name: Llama 3.2 3B modelSource: https://huggingface.co/meta-llama/Llama-3.2-3B offline: false privacy: private pricing: input: usd: 0.15 diem: 0.15 output: usd: 0.6 diem: 0.6 traits: - fastest object: model owned_by: venice.ai type: text StandardError: type: object properties: error: type: string description: A description of the error required: - error securitySchemes: BearerAuth: bearerFormat: JWT scheme: bearer type: http ````