SnapGen API

SnapGen is a prepaid AI gateway with OpenAI-compatible and Gemini-native text protocols plus image, video, and audio generation APIs. Change your base URL to https://api.snapgen.org/v1, keep your API key on the server, and select a model ID from the live catalog.

Quick start

Create an API key

Create an account, add balance, then create a key under Console > API Keys.

Choose a model

Call GET /v1/models or browse the model pages in the sidebar. Model availability is scoped to your key.

Send a request

Send the key as Authorization: Bearer sk-... and use the endpoint for the selected modality.

curl https://api.snapgen.org/v1/chat/completions \
  -H "Authorization: Bearer $SNAPGEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.4-mini",
    "messages": [{"role": "user", "content": "Reply with one sentence."}]
  }'

API surfaces

Text Series

GPT-5, Grok, Gemini, and DeepSeek chat models, billed per million tokens.

Image Series

GPT Image 2 and GPT Image 2.5 Ext through the OpenAI-compatible image endpoint.

Video Series

Seedance 2.0 and 2.5, Gemini Omni and Omni Flash 1.1, Veo 3.1, MiniMax H3, and Grok Imagine as asynchronous tasks.

Audio Series

ElevenLabs speech, transcription, music, sound effects, voice tools, and dubbing, plus Suno and Flow Music song generation and editing.

Development tools

Configure Cursor, Cline, Codex CLI, or Gemini CLI.

Text protocols

Each text model is served on the protocols listed on its page. SnapGen forwards a request only to routes that serve that protocol natively; it does not translate between protocols.

Protocol	Endpoint	Models
OpenAI Chat Completions	`POST /v1/chat/completions`	GPT-5, Grok, Gemini, DeepSeek
OpenAI Responses	`POST /v1/responses`	GPT-5, Grok
Gemini native	`POST /v1beta/models/{model}:generateContent`	Gemini
Anthropic Messages	`POST /v1/messages`	Claude (coming soon)

Audio endpoints

Speech and other ElevenLabs tools answer in the same HTTP response. Dubbing and every Suno and Flow Music model run as tasks that you read with GET /v1/tasks/{id}.

Endpoint	Returns	Models
`POST /v1/audio/speech`	Audio file	`eleven-v4`, `eleven-v3`, `eleven-multilingual-v2`, `eleven-flash-v2.5`
`POST /v1/audio/dialogue`	Audio file	`eleven-v4-dialogue`, `eleven-v3-dialogue`
`POST /v1/audio/transcriptions`	JSON transcript	`scribe-v2`
`POST /v1/audio/alignment`	JSON word timings	`eleven-forced-alignment`
`POST /v1/audio/music`	Audio file	`eleven-music-v2.5`
`POST /v1/audio/sound-effects`	Audio file	`eleven-sound-effects-v2`
`POST /v1/audio/voice-changer`	Audio file	`eleven-multilingual-sts-v2`, `eleven-english-sts-v2`
`POST /v1/audio/isolation`	Audio file	`eleven-voice-isolator`
`POST /v1/audio/dubbing`	Task; file from `GET /v1/tasks/{id}/content`	`eleven-dubbing-v1`
`POST /v1/music/generations`	Task; URLs in `result`	33 `suno-` and 9 `flow-music` models

Billing and availability

SnapGen uses prepaid USD balance with no subscription. Text models bill input, output, and cached tokens per million; GPT-5 requests whose prompt exceeds 272,000 tokens bill at the model’s long-context rate. Image models bill per delivered image. Video models bill per output second (by resolution) or per request. Audio models bill per character of text, per second of audio, or per request, as each model page lists. Failed generations are not charged. Catalog rates do not include payment-processor fees or promotions.

The live models page is the source of truth for current availability and pricing. Call GET /v1/models before hardcoding a model ID.

Overview

Text Series

Image Series

Video Series

Audio Series

Endpoint Reference

Task Management

Quick start

API surfaces

Text Series

Image Series

Video Series

Audio Series

Development tools

Text protocols

Audio endpoints

Billing and availability

​Quick start

​API surfaces

Text Series

Image Series

Video Series

Audio Series

Development tools

​Text protocols

​Audio endpoints

​Billing and availability

Quick start

API surfaces

Text protocols

Audio endpoints

Billing and availability