arivustudio_brief_from_url
ChatGPTRead a public web page and write an ad brief from it (one paragraph naming the product, brand, audience and what sets it apart), using only what the page says. Nothing is kept. Feed the brief to arivustudio_write_script. Costs very little from the Studio balance per brief.
arivustudio_cancel_generation
ChatGPTStop an image or video that is still queued or rendering, for good: it can't be resumed, and nothing is charged for it (starting it again is a new, charged generation). One that has already finished stays as it is, and cancelling twice changes nothing more.
arivustudio_change_voice
ChatGPTSay a recording again in another voice with ArivuStudio, keeping its words and timing. The voice is a standard voice from arivustudio_list_voices (engine standard), including the organization's cloned voices. The file is a finished item from an ArivuStudio library (fromGeneration) or a public link (url). The result is kept in the Voice Changer library and comes back with a link (valid 7 days). Paid from the Studio balance per minute, only when it is made; every call makes and charges a new file.
arivustudio_clone_voice
ChatGPTMake a new standard voice from a clear recording of one person speaking (10 to 60 seconds is best, up to 16 MB), for the organization's own use in speech, dubbing and the voice changer. Only with that person's explicit consent: ask the person you are helping whether the speaker agreed and who they are, and never clone a public figure or anyone who hasn't agreed. Without confirm: true it only says what it will do and starts nothing; call again with confirm: true only after the person confirms the speaker's consent in the chat. The file is a finished item from an ArivuStudio library (fromGeneration) or a public link (url). Paid from the Studio balance once per clone.
arivustudio_combine_videos
ChatGPTJoin 2 or more finished Image & Video videos, in the order given, into one new video on the first one's shape, each keeping its sound. The originals are kept. Free. It renders in the background: this returns the new video's id at once, then call arivustudio_get_generation with it to wait for the link.
arivustudio_dub
ChatGPTDub speech into another language with ArivuStudio: it transcribes the file, translates it and speaks the translation in a standard voice (arivustudio_list_voices with engine standard, including the organization's cloned voices). The file is a finished item from an ArivuStudio library (fromGeneration) or a public link (url). The dubbed audio is kept in the Dubbing library and comes back with a link (valid 7 days) and the translation. Paid from the Studio balance per minute transcribed plus per character spoken, only when it is made; every call makes and charges a new dub.
arivustudio_generate_image
ChatGPTMake new images with ArivuStudio from a text prompt, or a changed copy of an image made earlier (editOf; the original is kept). Waits up to about 40 seconds: finished images come back with a link (valid 7 days) and a small preview, and any still rendering come back as ids for arivustudio_get_generation. Each image costs its model's Studio credits, charged only when it finishes (prices: arivustudio_list_models). Every call starts new images and is charged again, so check on earlier ones with arivustudio_get_generation instead of calling this again.
arivustudio_generate_music
ChatGPTMake instrumental music from a description (English works best) with ArivuStudio. It is kept in the Music library and comes back with a link to the WAV (valid 7 days). Up to 30 seconds is newly composed and longer lengths loop it. Takes about 30 seconds. Spends Studio credits per track, only when it is made; every call makes and charges a new track.
arivustudio_generate_sound_effect
ChatGPTMake a sound effect from a description with ArivuStudio, 1 to 30 seconds. It is kept in the Sound Effects library and comes back with a link to the WAV (valid 7 days). It can take a minute or more: after about 40 seconds this answers that it is still being made, and it shows up in that library under its name (arivustudio_list_generations with tool "sound-effects"). Costs what the Omni Flash 1.1 clip it renders costs at the video rate (the shortest of 4, 6, 8 or 10 s that covers it; see arivustudio_list_models), charged only when it is made; every call makes and charges a new sound.
arivustudio_generate_video
ChatGPTStart new videos with ArivuStudio: from a text prompt, from an image made earlier (fromImage), or by continuing a video made earlier into a new, longer one (extendFrom; the original is kept). By default it makes one 8 second 720p video with the recommended model (Omni Flash 1.1). Without confirm: true it only says which model will run and about how many Studio credits it will spend, and starts nothing. Show that plan to the person and call again with confirm: true only after they agree to it in the chat, never in the same turn as the plan, and never for a different video than the one they asked for. Never show a real public figure, or a real person who hasn't agreed to it, even labelled as satire or fiction: decline and offer a made-up character instead. Videos take one to several minutes, so this returns their ids at once: call arivustudio_get_generation with them to wait for the result (a link valid 7 days). Credits are charged only for videos that finish. Every call with confirm: true starts new videos and is charged again, so don't repeat it to check on them.
arivustudio_get_credits
ChatGPTHow many ArivuStudio credits the organization has left and has used, and how much they still make at the defaults: images and videos on the recommended models (an 8 second video), characters of speech, music tracks and 5 second sound effects. available is what new work can spend now: the balance less what images and videos still rendering will be charged. Other models and lengths cost what arivustudio_list_models says.
arivustudio_get_generation
ChatGPTCheck on images and videos started with ArivuStudio, waiting up to about 40 seconds for them to finish (wait: false just looks). Finished ones come back with a link (valid 7 days; ask again for a fresh one), and images with a small preview. It also gives a fresh link to anything else in a library, such as speech, music or a sound effect. Read-only: it never starts or charges anything, so it is safe to call again.
arivustudio_isolate_voice
ChatGPTTake the background noise out of a recording with ArivuStudio, leaving the voice. The file is a finished item from an ArivuStudio library (fromGeneration) or a public link (url) to an audio or video file. The cleaned MP3 is kept in the Voice Isolator library and comes back with a link (valid 7 days). Free.
arivustudio_list_assets
ChatGPTList ArivuStudio's My Assets: only what people chose to save there (images, videos and audio, newest first, from the latest 200), each with a link (valid 7 days). What the tools make is kept in each tool's library instead: see arivustudio_list_generations.
arivustudio_list_generations
ChatGPTList what ArivuStudio has made, newest first. Every tool keeps what it makes in its own library: this reads one (Image & Video by default, or every library when asking for audio or text; speech, music, sound effects, transcripts and the rest by tool) or all of them at once. Each finished item comes with a link (valid 7 days). A transcript's prompt is its text, cut to 300 characters in this list: arivustudio_get_generation gives all of it. Cancelled ones are left out.
arivustudio_list_models
ChatGPTList the image and video models ArivuStudio can use, with each model's shapes (aspect ratios), sizes, resolutions, durations, whether a video model can start from an image or continue a video, and its price in Studio credits per output or per second. Pass a model's key to arivustudio_generate_image or arivustudio_generate_video.
arivustudio_list_voices
ChatGPTList voices for arivustudio_text_to_speech. engine "expressive" (default): 30 voices that act out tags such as [laughs] or [whispering] and can hold a dialogue. engine "standard": the voice library plus voices your organization cloned, each with a language.
arivustudio_text_to_speech
ChatGPTTurn text (up to 2000 characters) into speech with ArivuStudio. It is kept in the Text to Speech library and comes back with a link to the audio (valid 7 days). Expressive engine (default): one of its voices (default Kore), and tags such as [laughs], [whispering], [short pause], [pause] or [long pause] are performed, not read. For a dialogue pass 2 to 4 speakers and start every line of text with a speaker's name and a colon, e.g. "Asha: Hello!". Standard engine: a voice from arivustudio_list_voices with engine standard, and a language. Spends Studio credits per character, only when the audio is made; every call makes and charges a new file.
arivustudio_transcribe
ChatGPTTurn speech in an audio or video file into text with ArivuStudio, detecting the language. The file is a finished item from an ArivuStudio library (fromGeneration) or a public link (url). With speakers: true it also says who spoke when. The transcript is kept in the Speech to Text library. Paid from the Studio balance per minute of audio, only when it is transcribed; every call transcribes and charges again.
arivustudio_write_script
ChatGPTWrite a short voiceover script (about 30 to 60 words) from a brief with ArivuStudio, e.g. for an ad, or with kind "audiobook" rewrite text as narration. The script is kept in the Ads Engine library. Pair with arivustudio_brief_from_url to start from a web page, and arivustudio_text_to_speech to voice it. Costs very little from the Studio balance per script.