Models

Explore and browse every model available through ZnapAI.

`GET /v1/models`

Lists every model your API key can access.

curl --location 'https://api.znapai.com/v1/models' \
--header 'Authorization: Bearer $ZnapAI_API_KEY'

Response

{
  "data": [
    {
      "id": "gpt-image-2",
      "canonical_slug": "gpt-image-2",
      "name": "gpt-image-2",
      "created": 1677610602,
      "description": "",
      "context_window": null,
      "max_tokens": null,
      "mode": "image_generation",
      "architecture": {
        "modality": "text+image+file->image",
        "input_modalities": ["text", "image", "file"],
        "output_modalities": ["image"]
      },
      "links": {},
      "capabilities": {
        "supports_vision": true,
        "supports_pdf_input": true
      },
      "pricing": {
        "prompt": "0.0000025",
        "completion": "0.000005",
        "image": "0.000004",
        "image_output": "0.000015",
        "input_cache_read": "0.000000625"
      }
    },
    {
      "id": "gemini-3.1-flash-lite",
      "canonical_slug": "gemini-3.1-flash-lite",
      "name": "gemini-3.1-flash-lite",
      "created": 1677610602,
      "description": "",
      "context_window": 1048576,
      "max_tokens": 65536,
      "mode": "chat",
      "architecture": {
        "modality": "text+image+audio+file->text",
        "input_modalities": ["text", "image", "audio", "file"],
        "output_modalities": ["text"]
      },
      "links": {},
      "capabilities": {
        "supports_system_messages": true,
        "supports_response_schema": true,
        "supports_vision": true,
        "supports_function_calling": true,
        "supports_tool_choice": true,
        "supports_prompt_caching": true,
        "supports_audio_input": true,
        "supports_audio_output": false,
        "supports_pdf_input": true,
        "supports_native_streaming": true,
        "supports_web_search": true,
        "supports_url_context": true,
        "supports_reasoning": true
      },
      "pricing": {
        "prompt": "0.00000025",
        "completion": "0.0000015",
        "audio": "0.0000005",
        "internal_reasoning": "0.0000015",
        "input_cache_read": "0.000000025",
        "web_search": "0.014"
      }
    }
    /* ...remaining models */
  ],
  "total_count": 61,
  "links": {
    "next": null
  }
}

Query Parameters

The Models API supports query parameters to filter the list of models returned.

`mode`

Filter models by what they’re used for — a comma-separated list.

Value	Description
`chat`	Text and multimodal chat/completion models
`image_generation`	Models that generate images
`video_generation`	Models that generate video
`embedding`	Embedding models
`rerank`	Reranking models
`responses`	Models served through the Responses API
`realtime`	Realtime (low-latency voice/streaming) models
`audio_speech`	Text-to-speech models
`audio_transcription`	Speech-to-text models

curl --location 'https://api.znapai.com/v1/models?mode=video_generation' \
--header 'Authorization: Bearer $ZnapAI_API_KEY'

Response

{
  "data": [
    {
      "id": "veo-3.1-generate-preview",
      "canonical_slug": "veo-3.1-generate-preview",
      "name": "veo-3.1-generate-preview",
      "created": 1677610602,
      "description": "",
      "context_window": 1024,
      "max_tokens": null,
      "mode": "video_generation",
      "architecture": {
        "modality": "text->video",
        "input_modalities": ["text"],
        "output_modalities": ["video"]
      },
      "links": {},
      "pricing": {
        "prompt": null,
        "completion": null,
        "request": null,
        "image": null,
        "image_output": null,
        "audio": null,
        "audio_output": null,
        "internal_reasoning": null,
        "input_cache_read": null,
        "input_cache_write": null,
        "input_cache_write_1h": null,
        "web_search": null
      }
    },
    {
      "id": "veo-3.1-fast-generate-preview",
      "canonical_slug": "veo-3.1-fast-generate-preview",
      "name": "veo-3.1-fast-generate-preview",
      "created": 1677610602,
      "description": "",
      "context_window": 1024,
      "max_tokens": null,
      "mode": "video_generation",
      "architecture": {
        "modality": "text->video",
        "input_modalities": ["text"],
        "output_modalities": ["video"]
      },
      "links": {},
      "pricing": {
        "prompt": null,
        "completion": null,
        "request": null,
        "image": null,
        "image_output": null,
        "audio": null,
        "audio_output": null,
        "internal_reasoning": null,
        "input_cache_read": null,
        "input_cache_write": null,
        "input_cache_write_1h": null,
        "web_search": null
      }
    }
    /* ...remaining video models */
  ],
  "total_count": 6,
  "links": {
    "next": null
  }
}

`output_modalities`

Filter models by their output capabilities. Accepts a comma-separated list of modalities.

Value	Description
`text`	Models that produce text output
`image`	Models that generate images
`audio`	Models that produce audio output
`embeddings`	Embedding models
`video`	Models that generate video

Examples:

# All models (default)
curl --location 'https://api.znapai.com/v1/models' \
--header 'Authorization: Bearer $ZnapAI_API_KEY'

# Image generation models only
curl --location 'https://api.znapai.com/v1/models?output_modalities=image' \
--header 'Authorization: Bearer $ZnapAI_API_KEY'

# Text and image models
curl --location 'https://api.znapai.com/v1/models?output_modalities=text,image' \
--header 'Authorization: Bearer $ZnapAI_API_KEY'

Response

{
  "data": [
    {
      "id": "gpt-image-2",
      "canonical_slug": "gpt-image-2",
      "name": "gpt-image-2",
      "created": 1677610602,
      "description": "",
      "context_window": null,
      "max_tokens": null,
      "mode": "image_generation",
      "architecture": {
        "modality": "text+image+file->image",
        "input_modalities": ["text", "image", "file"],
        "output_modalities": ["image"]
      },
      "links": {},
      "capabilities": {
        "supports_vision": true,
        "supports_pdf_input": true
      },
      "pricing": {
        "prompt": "0.0000025",
        "completion": "0.000005",
        "image": "0.000004",
        "image_output": "0.000015",
        "input_cache_read": "0.000000625"
      }
    },
    {
      "id": "gpt-5.2",
      "canonical_slug": "gpt-5.2",
      "name": "gpt-5.2",
      "created": 1677610602,
      "description": "",
      "context_window": 272000,
      "max_tokens": 128000,
      "mode": "chat",
      "architecture": {
        "modality": "text+image->text+image",
        "input_modalities": ["text", "image"],
        "output_modalities": ["text", "image"]
      },
      "links": {},
      "capabilities": {
        "supports_system_messages": true,
        "supports_response_schema": true,
        "supports_vision": true,
        "supports_function_calling": true
        /* ...remaining capabilities */
      },
      "pricing": {
        "prompt": "0.0000021",
        "completion": "0.0000084",
        "input_cache_read": "0.00000021"
      }
    }
    /* ...remaining image-output models */
  ],
  "total_count": 5,
  "links": {
    "next": null
  }
}

Model Object Schema

Each model in the data array contains the following fields:

Field	Type	Description
`id`	`string`	The model name you pass in API requests (e.g. `"gemini-3.1-flash-lite"`)
`canonical_slug`	`string`	Always identical to `id` on ZnapAI
`name`	`string`	Human-readable display name
`created`	`number`	Unix timestamp of when the model was added
`description`	`string`	A short description of the model, when available
`context_window`	`number \| null`	Maximum number of input tokens the model accepts
`max_tokens`	`number \| null`	Maximum number of output tokens the model can generate
`mode`	`string \| null`	What the model is used for — `chat`, `image_generation`, `video_generation`, `embedding`, `rerank`, `audio_transcription`, etc.
`architecture`	`Architecture`	Object describing the model’s input/output modalities
`links`	`object`	Reserved for future use
`capabilities`	`Capabilities \| undefined`	The model’s supported features (omitted if none are known)
`pricing`	`Pricing \| undefined`	What you pay to use the model, in USD per token (omitted if pricing isn’t available) — see Pricing

Architecture Object

{
  "modality": string,           // e.g. "text+image->text"
  "input_modalities": string[], // e.g. ["text", "image"]
  "output_modalities": string[] // e.g. ["text"]
}

input_modalities — the types of input the model accepts (text, image, audio, file).
output_modalities — the types of output the model can produce.

Capabilities Object

The model’s supported features, when known — things like supports_vision, supports_function_calling, supports_reasoning, and, for reasoning models, which reasoning effort levels they support (supports_none_reasoning_effort, supports_xhigh_reasoning_effort, etc.). Omitted entirely if no capability data is available for a model.

Pricing

pricing shows what you actually pay to use the model — already reflecting any discounts, in USD per token (or per request, for models like rerank).

Fields

All values are USD amounts per token, formatted as strings:

Field	Description
`prompt`	Cost per input (prompt) token
`completion`	Cost per output (completion) token
`request`	Flat cost per request (used by models like rerank instead of per-token pricing)
`image`	Cost per input image token
`image_output`	Cost per output image token
`audio`	Cost per input audio token
`audio_output`	Cost per output audio token
`internal_reasoning`	Cost per reasoning token
`input_cache_read`	Cost per cached input token that’s read
`input_cache_write`	Cost per cached input token that’s written
`input_cache_write_1h`	Cost per cached input token written with a 1-hour cache lifetime
`web_search`	Cost per web search performed

Fields only appear when they’re relevant to a model — for example, image-output models show image_output, audio-capable models show audio/audio_output, and text-only models show neither. For most models, pricing is omitted entirely if pricing isn’t available.

Video and rerank models

For video generation and rerank models, pricing always appears, with null for any field that doesn’t apply — since these models are often priced in ways (e.g. per-second video) that don’t fit the standard per-token fields above.

`overrides` (long-context pricing)

Some models charge more once a request passes a certain prompt length. When that applies, overrides lists the higher rates and the token threshold they kick in at:

"overrides": [
  { "min_prompt_tokens": 128000, "prompt": "...", "completion": "..." },
  { "min_prompt_tokens": 200000, "prompt": "...", "completion": "...", "input_cache_read": "...", "input_cache_write": "..." }
]

Omitted entirely for models without long-context pricing.

Example response

{
  "id": "gemini-3.1-flash-lite",
  "canonical_slug": "gemini-3.1-flash-lite",
  "name": "gemini-3.1-flash-lite",
  "created": 1677610602,
  "description": "",
  "context_window": 1048576,
  "max_tokens": 65536,
  "mode": "chat",
  "architecture": {
    "modality": "text+image+audio+file->text",
    "input_modalities": ["text", "image", "audio", "file"],
    "output_modalities": ["text"]
  },
  "links": {},
  "capabilities": {
    "supports_system_messages": true,
    "supports_response_schema": true,
    "supports_vision": true,
    "supports_function_calling": true,
    "supports_tool_choice": true,
    "supports_prompt_caching": true,
    "supports_audio_input": true,
    "supports_audio_output": false,
    "supports_pdf_input": true,
    "supports_native_streaming": true,
    "supports_web_search": true,
    "supports_url_context": true,
    "supports_reasoning": true
  },
  "pricing": {
    "prompt": "0.00000025",
    "completion": "0.0000015",
    "audio": "0.0000005",
    "internal_reasoning": "0.0000015",
    "input_cache_read": "0.000000025",
    "web_search": "0.014"
  }
}

A video-generation model, for comparison — note capabilities is omitted entirely when no capability data is available for it, while pricing still appears with null fields:

{
  "id": "veo-3.1-generate-preview",
  "canonical_slug": "veo-3.1-generate-preview",
  "name": "veo-3.1-generate-preview",
  "created": 1677610602,
  "description": "",
  "context_window": 1024,
  "max_tokens": null,
  "mode": "video_generation",
  "architecture": {
    "modality": "text->video",
    "input_modalities": ["text"],
    "output_modalities": ["video"]
  },
  "links": {},
  "pricing": {
    "prompt": null,
    "completion": null,
    "request": null,
    "image": null,
    "image_output": null,
    "audio": null,
    "audio_output": null,
    "internal_reasoning": null,
    "input_cache_read": null,
    "input_cache_write": null,
    "input_cache_write_1h": null,
    "web_search": null
  }
}

`GET /models` (OpenAI format)

Dropping the v1 prefix returns a lighter, OpenAI-compatible model list — just the id, object, created, and owned_by fields, matching the shape returned by OpenAI’s own /models endpoint. Use this if you’re pointing an existing OpenAI SDK or tool at ZnapAI and just need the standard model list.

curl --location 'https://api.znapai.com/models' \
--header 'Authorization: Bearer $ZnapAI_API_KEY'

Response

{
  "data": [
    {
      "id": "gpt-image-2",
      "object": "model",
      "created": 1677610602,
      "owned_by": "openai"
    },
    {
      "id": "gemini-3.1-flash-lite",
      "object": "model",
      "created": 1677610602,
      "owned_by": "openai"
    },
    {
      "id": "veo-3.1-generate-preview",
      "object": "model",
      "created": 1677610602,
      "owned_by": "openai"
    }
    /* ...remaining models */
  ]
}

Usage

Integrations

Models

`GET /v1/models`

Query Parameters

`mode`

`output_modalities`

Model Object Schema

Architecture Object

Capabilities Object

Pricing

Fields

Video and rerank models

`overrides` (long-context pricing)

Example response

`GET /models` (OpenAI format)

​GET /v1/models

​Query Parameters

​mode

​output_modalities

​Model Object Schema

​Architecture Object

​Capabilities Object

​Pricing

​Fields

​Video and rerank models

​overrides (long-context pricing)

​Example response

​GET /models (OpenAI format)

`GET /v1/models`

Query Parameters

`mode`

`output_modalities`

Model Object Schema

Architecture Object

Capabilities Object

Pricing

Fields

Video and rerank models

`overrides` (long-context pricing)

Example response

`GET /models` (OpenAI format)