> ## Documentation Index > Fetch the complete documentation index at: https://docs.apostra.com/llms.txt > Use this file to discover all available pages before exploring further. # AudioStack creative-tool compatibility > Current public support and limitations for AudioStack audio generation in Apostra AudioStack is a creative-only audio generation service. It is not a direct media-buy adapter. Apostra uses it to turn a script and voice settings into one rendered `audio_hosted` creative manifest. **Last reviewed 2026-09-13.** This page describes what Apostra integration offers. AudioStack has not reviewed or endorsed it. Availability can still vary by account, voice, language, and provider access. ## Roadmap The AudioStack integration roadmap is tracked as AI-6601 (a plain-text reference — Apostra's internal issue tracker is not public). The issue covers creative discovery, reusable-library and multi-output gaps, and live validation. The issue key remains a stable reference even for readers who cannot access the internal Linear workspace. ## Connection, accounts, and identity AudioStack connects to Apostra through an API key, not OAuth: the key is an access grant billed to one authorized organization. There is no ad account or manager-account hierarchy to discover. The advertiser you map to that connection chooses which AudioStack organization is billed for a generation; your brand's `voice_synthesis` setting chooses the voice. Mapping and fail-closed behavior otherwise follow the same model as ad platforms — see [Platform accounts and advertiser mapping](/v2/concepts/ad-platform-accounts). AudioStack does not yet have a dedicated section in the connecting guide. ## What you can do through Apostra | Feature | Availability | References | Comments | | ---------------------------------------------- | ------------ | ----------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Build an audio creative (`build_creative`) | 🟢 Available | [AudioStack API overview](https://docs.audiostack.ai/docs/audiostack-api) | One request performs a full AudioStack render and returns one `audio_hosted` creative containing the rendered audio. | | Preview without rendering (`preview_creative`) | ⚪ Planned | [AudioStack API overview](https://docs.audiostack.ai/docs/audiostack-api) · AI-6601 | To audition an output, run a full `build_creative` render and play the returned audio; there is no preview-only path. Current boundary: Apostra implementation/testing. | | Reuse an AudioStack library creative | ⚪ Planned | [AudioStack API overview](https://docs.audiostack.ai/docs/audiostack-api) · AI-6601 | The integration does not look up, sync, list, or read back creatives in an AudioStack library. Each build starts a new render. Current boundary: Apostra implementation/testing. | | Produce several lengths in one request | ⚪ Planned | [AudioStack API overview](https://docs.audiostack.ai/docs/audiostack-api) · AI-6601 | One build selects one duration. Submit separate builds for 15-second and 30-second versions. Current boundary: Apostra implementation/testing. | | Buy media | ⚪ Planned | — | AudioStack is used for creative generation, not products, campaigns, targeting, or delivery reporting. This is outside the integration's scope rather than a gap on the roadmap above. | ## Supported build shape AudioStack accepts the canonical `audio_hosted` format. A build requires a text `script` asset and can bind a voice, language, brand name, speed, and music template. The current runtime supports one of these output choices per call: * `duration_ms_exact: 15000` for a 15-second audio ad; * `duration_ms_exact: 30000` for a 30-second audio ad; or * no exact duration for the voiceover path. Voiceovers use AudioStack's Audioform service at the script's natural spoken length, without background music. They do not use the deprecated script and speech endpoints. Legacy agent-URL format references are not accepted. AudioStack does not publish a separate buyer-selectable format catalog; `list_creative_formats` returns an empty list because the integration consumes canonical `audio_hosted` directly. ## Webhooks and async lifecycle Not applicable. `build_creative` is synchronous: one request performs the full render and returns the result, so there is no async task and no webhook/polling distinction to make here — unlike the direct ad-platform adapters, which use polling (see each adapter's own "Webhooks and async lifecycle" section). ## Reporting Not applicable. AudioStack is a creative-only integration; it does not report delivery, spend, or performance metrics. See [Reporting overview](/v2/guides/reporting-overview) for how ad-platform delivery reporting works. ## Provider and account constraints AudioStack access is bring-your-own-key. The connected credential resolves one billable AudioStack organization, and the advertiser must be mapped to that provider connection before generation. Authentication, provider rate limits, voice availability, and upstream rendering availability can still block an otherwise valid build. Apostra does not publish a market or voice availability list for AudioStack; the selected provider account determines what is available at request time. ## Provider documentation * [AudioStack API overview](https://docs.audiostack.ai/docs/audiostack-api) * [Generative creative provider credentials](/v2/buyer/creatives/generative-creative)