# Albastar API

> Developer documentation for the Albastar API at https://api.albastov.com/v1, for albastov.com accounts and their keys: chat completions, text to speech, voices and voice cloning, speech to text, images, and events. Every link below is the page as plain markdown; the rendered pages are at https://docs.albastov.com.

Send the key as `Authorization: Bearer <key>`. The same pages as JSON: `GET https://api.albastov.com/v1/docs` (index) and `GET https://api.albastov.com/v1/docs/{section}/{page}`.

Apps in the browser, for albastov.com accounts: [GenAI](https://genai.albastov.com), images and videos from a prompt; [Music](https://music.albastov.com), songs with vocals or instrumentals (no documented API routes yet).

## Get started

- [Documentation](https://api.albastov.com/v1/docs/get-started/introduction.md): Language models, speech, voices and events at api.albastov.com.
- [Quick start](https://api.albastov.com/v1/docs/get-started/quick-start.md): Get a key, send a chat completion, read the reply and make a speech file.

## Language models

- [Language models](https://api.albastov.com/v1/docs/language-models/overview.md): OpenAI-compatible chat completions on open-weight Qwen and DeepSeek models.
- [Chat completions](https://api.albastov.com/v1/docs/language-models/chat-completions.md): Messages, models, conversation history, OpenAI SDKs, prompt caching, model loading, retries and limits on POST /v1/chat/completions.
- [Streaming](https://api.albastov.com/v1/docs/language-models/streaming.md): Chat completions as server-sent events, token by token.
- [Tools and structured output](https://api.albastov.com/v1/docs/language-models/tools-and-structured-output.md): Function calling with tools, and JSON output that matches a schema.
- [Thinking](https://api.albastov.com/v1/docs/language-models/thinking.md): Reasoning before the answer, on by default, off with chat_template_kwargs.enable_thinking.

## Audio

- [Audio](https://api.albastov.com/v1/docs/audio/overview.md): Text to speech in built-in, cloned and designed voices, and speech to text with word timestamps.
- [Text to speech](https://api.albastov.com/v1/docs/audio/text-to-speech.md): Speech from text with POST /v1/audio/speech, in a built-in voice or a saved voice.
- [Voices](https://api.albastov.com/v1/docs/audio/voices.md): Clone, design, list, copy and delete saved voices, and speak any text in a saved voice.
- [Voice cloning](https://api.albastov.com/v1/docs/audio/voice-cloning.md): Speech in the voice of a reference clip on 12 cloning models, with the clip rules and consent.
- [Speech to text](https://api.albastov.com/v1/docs/audio/speech-to-text.md): Transcripts of audio files with POST /v1/audio/transcriptions, with word and segment timestamps.

## Image

- [Image](https://api.albastov.com/v1/docs/image/overview.md): Images from a prompt or from a starting image, as media jobs with POST /v1/media/jobs.

## Music

- [Music](https://api.albastov.com/v1/docs/music/overview.md): Songs in the music studio at music.albastov.com; the Albastar API documents no music routes yet.

## Events

- [Events](https://api.albastov.com/v1/docs/events/overview.md): One-line notices from your apps, stored per source and streamed to readers.
- [Publishing](https://api.albastov.com/v1/docs/events/publishing.md): Store a publish key, name the source, publish events from an app, and set each event's fields and lifetime.
- [Webhooks](https://api.albastov.com/v1/docs/events/webhooks.md): Send another app's webhooks to a webhook URL that Albastar Corp. sets up, and read each delivery's outcome.
- [Receiving](https://api.albastov.com/v1/docs/events/receiving.md): Read keys, listing, the live stream, reconnecting, and dismissal.

## Models

- [Chat models](https://api.albastov.com/v1/docs/models/chat-models.md): Open-weight Qwen and DeepSeek chat models on one OpenAI-compatible route.
- [DeepSeek V4.1 Flash](https://api.albastov.com/v1/docs/models/deepseek-v4-1-flash.md): Mixture-of-experts model; 8,192-token context for prompt, reasoning and answer together, one request at a time.
- [Qwen3.8 27B](https://api.albastov.com/v1/docs/models/qwen3-8-27b.md): 150,000-token context, 64 parallel requests per account, tools and JSON output.
- [Qwen3.8 27B Fast](https://api.albastov.com/v1/docs/models/qwen3-8-27b-fast.md): 65,536-token context, 8 parallel requests per account, tools and JSON output.
- [Qwen3.8 27B Uncensored](https://api.albastov.com/v1/docs/models/qwen3-8-27b-uncensored.md): Refusal behaviour removed; 150,000-token context, 64 parallel requests per account, tools and JSON output.
- [Qwen3.8 Flash-Next Abliterated](https://api.albastov.com/v1/docs/models/qwen3-8-flash-next-abliterated.md): Mixture-of-experts model with refusal behaviour removed; 131,072-token context, one request at a time.
- [Voice models](https://api.albastov.com/v1/docs/models/voice-models.md): Text-to-speech models by name, with the capabilities, languages and a sample of each.
- [BreezeTTS 2](https://api.albastov.com/v1/docs/models/breeze2.md): Clones a voice in Chinese and English; `instruct` describes the delivery.
- [Chatterbox](https://api.albastov.com/v1/docs/models/chatterbox.md): Clones a voice in 19 languages, using up to 20 s of the clip.
- [CosyVoice 3](https://api.albastov.com/v1/docs/models/cosyvoice3.md): Clones a voice in 9 languages; the clip and the text may be in different languages.
- [FireRedTTS 3](https://api.albastov.com/v1/docs/models/fireredtts3.md): Clones a voice in 24 languages, including Arabic, Thai and Vietnamese.
- [Fish Audio S2 Pro](https://api.albastov.com/v1/docs/models/fish-s2pro.md): Clones a voice in 80+ languages, using up to 20 s of the clip.
- [Higgs Audio v3](https://api.albastov.com/v1/docs/models/higgs3.md): Clones a voice in 100+ languages, using up to 20 s of the clip.
- [IndexTTS 2.5](https://api.albastov.com/v1/docs/models/indextts2-5.md): Clones a voice in English or Chinese; `language` on a media job selects Chinese.
- [Kokoro](https://api.albastov.com/v1/docs/models/kokoro.md): Speaks text in 54 built-in voices in 8 languages; `speed` sets the rate from 0.5 to 2.0.
- [MioTTS](https://api.albastov.com/v1/docs/models/miotts.md): Clones a voice in English and Japanese, using up to 20 s of the clip.
- [MOSS-TTS Local Transformer](https://api.albastov.com/v1/docs/models/moss-local.md): Clones a voice in 31 languages, including Cantonese, Persian and Tagalog.
- [OmniVoice](https://api.albastov.com/v1/docs/models/omnivoice.md): Clones a voice, or designs one from a description in `instruct`, in 600+ languages.
- [Qwen3-TTS](https://api.albastov.com/v1/docs/models/qwen3-tts.md): Clones a voice in 10 languages, using up to 22 s of the clip.
- [VoxCPM 2](https://api.albastov.com/v1/docs/models/voxcpm2.md): Clones a voice in 30 languages, detected from the text, using up to 11 s of the clip.
- [Transcription models](https://api.albastov.com/v1/docs/models/transcription-models.md): Speech-to-text models for POST /v1/audio/transcriptions, with the capabilities and languages of each.
- [Parakeet](https://api.albastov.com/v1/docs/models/parakeet.md): Transcribes speech in 25 European languages, with punctuation and optional word timestamps.

## API reference

- [API reference](https://api.albastov.com/v1/docs/api-reference/overview.md): Every route of the Albastar API, with its parameters, responses and errors.
- [Authentication](https://api.albastov.com/v1/docs/api-reference/authentication.md): API keys, their scopes and the Authorization header.
- [Errors](https://api.albastov.com/v1/docs/api-reference/errors.md): Error shapes, status codes, key errors and retry rules.
- [Rate limits](https://api.albastov.com/v1/docs/api-reference/rate-limits.md): Call, token, size and wait limits, with the error each returns.
- [Chat completions](https://api.albastov.com/v1/docs/api-reference/chat-completions.md): POST /v1/chat/completions parameters, headers, response and errors.
- [List models](https://api.albastov.com/v1/docs/api-reference/models-list.md): GET /v1/models response fields and errors.
- [Speech](https://api.albastov.com/v1/docs/api-reference/speech.md): POST /v1/audio/speech returns an audio file of the text in a built-in or saved voice.
- [Transcriptions](https://api.albastov.com/v1/docs/api-reference/transcriptions.md): POST /v1/audio/transcriptions returns the transcript of an audio file, with optional timestamps.
- [Voices](https://api.albastov.com/v1/docs/api-reference/voices.md): List, create, design, read, change, copy and delete saved voices, and fetch their clips.
- [Media jobs](https://api.albastov.com/v1/docs/api-reference/media-jobs.md): Speech, transcription and image jobs, the wait for a result, queue timeouts, and the model list with parameters.
- [Uploads](https://api.albastov.com/v1/docs/api-reference/uploads.md): POST /v1/media/uploads stores an audio or image file on the account and returns its id up_….
- [Samples](https://api.albastov.com/v1/docs/api-reference/samples.md): GET /v1/media/samples lists the reference clip and one sample clip per text-to-speech model; no key needed.
- [Events](https://api.albastov.com/v1/docs/api-reference/events.md): Publish events, receive webhooks, and list, stream and dismiss events.
- [Docs](https://api.albastov.com/v1/docs/api-reference/docs.md): The documentation as JSON, as markdown and as an llms.txt index, and its images, without a key.
- [Status](https://api.albastov.com/v1/docs/api-reference/status.md): GET /v1/status, the current state of the API and of each product, with 90 days of daily states and uptime.

## Changelog

- [Changelog](https://api.albastov.com/v1/docs/changelog/changelog.md): Every change to the Albastar API, newest first.