SpicyAPI

What it is

SpicyAPI is an API platform that aggregates 193 AI models across image, video and text generation tasks, offering both mainstream and uncensored model editions. The platform provides a single API key and unified billing in US dollars with no subscription, allowing developers to access models from publishers including Alibaba, ByteDance, OpenAI, Anthropic, and others.

The platform operates two interfaces: Spicy Studio for creators to generate images and videos without code, and a developer API supporting OpenAI-compatible formats for chat, REST endpoints for media tasks, and integration with AI agents via MCP. Each model is individually integrated with transparent pricing per unit, content policy declarations, and straightforward failure handling through idempotency keys and signed callbacks.

SpicyAPI emphasizes operational transparency and data control, with user-configurable retention periods for uploads, generated media and prompts, per-key spend caps with daily and lifetime limits, and explicit statements about not training models on user data. The platform runs 193 model families across 21 task types including text-to-video, image editing, character animation and chat.

Questions from the bench

Is SpicyAPI an uncensored AI API?

Yes, at the platform level SpicyAPI adds no content classifier, no per-request declaration and no approval queue. Output depends on the model chosen—Spicy editions are built for unrestricted generation, while mainstream models keep their own rules. Each model page states what it accepts.

How much does a generation cost?

Prices are in US dollars, quoted per model in their own unit—per image, per second of video or per million tokens. The Playground shows the estimate for exact settings before submission, and active promotions show original and discounted prices side by side.

Is the API OpenAI-compatible?

For chat, yes—point any OpenAI client at SpicyAPI's base URL, swap in your key and model name, and streaming works as-is. Video also offers the /v1/videos endpoint in OpenAI Videos format. Image and video tasks use createTask with recordInfo polling or a callBackUrl.

Can I build an AI companion or roleplay app on SpicyAPI?

Yes. Pair an OpenAI-compatible LLM for in-character dialogue with image and video models for portraits and short clips, all on one key and balance. To keep character consistent, reuse the same reference image as first frame or reference input for each shot.

Which models support uncensored generation?

Start with the Spicy editions — the model name ends in “Spicy” — and check the catalogue for the tasks each one covers. Several mainstream image and video models also follow uncensored prompts without adding a filter of their own; every model page states what it accepts. For image-to-video, begin from a still you like: the first frame steers the character and composition.