Media model reference

These are the underlying models the generate_image / generate_video / generate_audio verbs resolve to. You don't call them directly — you pick a model family with a verb and the platform adapts your inputs to the native shape below. This page documents that native shape for reference.

For the verbs, see agent-tools-reference and sdk-runtime-reference. For prices, see the pricing page.

Index

Images

openai/gpt-image-2

GPT Image 2 · text → image

Quality × size matrix. Set quality ∈ low|medium|high and size (e.g. 1024x1024).

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

openai/gpt-image-2-edit

GPT Image 2 (edit) · image → image (edit)

Image-to-image edit. Quality × size matrix; provide image_url(s).

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

bytedance/seedream-v4

Seedream v4 · text → image

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

bytedance/seedream-v4-edit

Seedream v4 (edit) · image → image (edit)

Edit/composite reference images. Provide image_url (or list) plus a prompt.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

google/imagen-4

Imagen 4 · text → image

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/nano-banana-pro

Nano Banana Pro · text → image

Reasoning image gen with strong multi-language text rendering and native up-to-4K output. Set resolution ∈ 1K|2K|4K.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/nano-banana-pro-edit

Nano Banana Pro (edit) · image → image (edit)

Edit/compose from up to 14 reference images via image_urls + a prompt. Strong text rendering, native up-to-4K.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

kuaishou/kling-v3-image

Kling v3 (image) · text → image

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

kuaishou/kling-v3-image-edit

Kling v3 (image edit) · image → image (edit)

Edit/compose from a base reference image (+ optional extra references). Provide refs.

Inputs

Max file size: 10.0MB, Min width: 300px, Min height: 300px, Min aspect ratio: 0.40, Max aspect ratio: 2.50, Timeout: 20.0s | | prompt | string | ✓ | — | Text prompt for image generation. Max 2500 characters. (max 2,500 chars) | | aspect_ratio | enum | | "16:9" | Aspect ratio of generated images. Values: "16:9" | "9:16" | "1:1" | "4:3" | "3:4" | "3:2" | "2:3" | "21:9" | | elements | array<elementinput> | | — | Optional: Elements (characters/objects) to include in the image for face control | | num_images | integer | | 1 | Number of images to generate (1-9). (≥ 1, ≤ 9) | | output_format | enum | | "png" | The format of the generated image. Values: "jpeg" | "png" | "webp" | | resolution | enum | | "1K" | Image generation resolution. 1K: standard, 2K: high-res. Values: "1K" | "2K" |

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

Videos

bytedance/seedance-2-t2v

Seedance 2.0 — text→video

720p–1080p text-to-video. Audio included. Per-second rate jumps at 1080p.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

bytedance/seedance-2-i2v

Seedance 2.0 — image→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

bytedance/seedance-2-r2v

Seedance 2.0 — reference→video

Up to 9 image / 3 video / 3 audio references. Per-second drops 40% when a video reference is passed.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

bytedance/seedance-2-fast-t2v

Seedance 2.0 Fast — text→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

bytedance/seedance-2-fast-i2v

Seedance 2.0 Fast — image→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

bytedance/seedance-2-fast-r2v

Seedance 2.0 Fast — reference→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one). Additional metadata available under meta (seed).

kuaishou/kling-v3-t2v

Kling v3 Pro — text→video

Set generate_audio: true to enable audio, voice_control: true for voice.

Inputs

        the model to stick to your prompt. (≥ 0, ≤ 1) |

| duration | enum | | "5" | The duration of the generated video in seconds. Values: "3" | "4" | "5" | "6" | "7" | "8" | "9" | "10" | "11" | "12" | "13" | "14" | "15" | | generate_audio | boolean | | true | Whether to generate native audio for the video. Supports Chinese and English voice output. Other languages are automatically translated to English. For English speech, use lowercase letters; for acronyms or proper nouns, use uppercase | | multi_prompt | array<klingv3multipromptelement> | | — | List of prompts for multi-shot video generation. If provided, overrides the single prompt and divides the video into multiple shots with specified prompts and durations | | negative_prompt | string | | "blur, distort, and low quality" | (max 2,500 chars) | | prompt | string | | — | Text prompt for video generation. Either prompt or multi_prompt must be provided, but not both | | shot_type | enum | | "customize" | The type of multi-shot video generation. 'intelligent' lets the model automatically determine shot structure. Values: "customize" | "intelligent" |

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

kuaishou/kling-v3-i2v

Kling v3 Pro — image→video

Inputs

        the model to stick to your prompt. (≥ 0, ≤ 1) |

| duration | enum | | "5" | The duration of the generated video in seconds. Values: "3" | "4" | "5" | "6" | "7" | "8" | "9" | "10" | "11" | "12" | "13" | "14" | "15" | | elements | array<klingv3comboelementinput> | | — | Elements (characters/objects) to include in the video. Each example can either be an image set (frontal + reference images) or a video. Reference in prompt as @Element1, @Element2, etc | | end_image_url | string | | — | URL of the image to be used for the end of the video | | generate_audio | boolean | | true | Whether to generate native audio for the video. Supports Chinese and English voice output. Other languages are automatically translated to English. For English speech, use lowercase letters; for acronyms or proper nouns, use uppercase | | multi_prompt | array<klingv3multipromptelement> | | — | List of prompts for multi-shot video generation. If provided, divides the video into multiple shots | | negative_prompt | string | | "blur, distort, and low quality" | (max 2,500 chars) | | prompt | string | | — | Text prompt for video generation. Either prompt or multi_prompt must be provided, but not both | | shot_type | enum | | "customize" | The type of multi-shot video generation. 'intelligent' lets the model automatically determine shot structure. Values: "customize" | "intelligent" |

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

kuaishou/kling-o3-r2v

Kling O3 Pro — reference→video

Reference-to-video on the Kling O3 line. Pass refs as image_urls; default 8s.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

kuaishou/kling-avatar-v2

Kling AI Avatar v2 Pro — lip-sync · video

Lip-synced talking-head from a portrait image_url + an audio_url. Output duration auto-matches the audio; framing follows the image. Optional prompt for delivery/expression. Billed per second of output.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/veo-3-t2v

Veo 3 — text→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/veo-3-i2v

Veo 3 — image→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/veo-3-fast-t2v

Veo 3 Fast — text→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/veo-3-fast-i2v

Veo 3 Fast — image→video

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

google/veo-3.1-r2v

Veo 3.1 — reference→video

Reference-to-video on Veo 3.1. Pass reference images as image_urls.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

Audio

elevenlabs/scribe-v2

ElevenLabs Scribe v2 — speech→text

Speech-to-text with word-level timestamps. Pass audio_url; the call returns text + words (each with start/end seconds) instead of a file. Pass keyterms (brand / proper nouns) to bias the transcription. Billed per second of audio.

Inputs

Output

Returns the transcript inline as text plus a words array (each word carries start/end seconds and type) and a detected language_code. No file is produced, so drive_path is empty.

elevenlabs/tts-v3

ElevenLabs v3 — expressive text→speech

Expressive multilingual text-to-speech (74 languages) that reads inline audio tags — wrap a cue in square brackets ([excited], [whispers], [laughs], [sighs], [sarcastic], [British accent]) and v3 shapes the read accordingly; the tag is acted, never spoken. Pass text, an optional voice (preset name like "Aria"/"Roger" or a voice id), and stability (0.0 = most expressive and tag-responsive, 0.5 = balanced, 1.0 = steadiest/least tag-responsive). v3 tunes delivery through tags + stability, so it ignores the style/speed/similarity_boost knobs — pick elevenlabs/tts-multilingual-v2 when you need those instead. Billed per character of text.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).

elevenlabs/tts-multilingual-v2

ElevenLabs Multilingual v2 — text→speech

Natural multilingual text-to-speech (29 languages) with fine delivery controls — stability, style, speed, similarity_boost. Does not interpret audio tags (use elevenlabs/tts-v3 for [excited]/[whispers]-style direction). Pass text (and an optional voice). Billed per character of text.

Inputs

Output

Saved to your project drive at drive_path (mint a URL with puras.drive.url / the drive_url tool if you need one).