Does FastGen Cloud rewrite my prompt like revised_prompt does?

No. There is no language model in front of the sampler. The only strings prepended are a checkpoint’s native quality tags and a LoRA trigger word when you select a catalog LoRA, both documented on the model page.

Is there a moderation setting I can turn down?

There is no dial, because there is no general content filter to turn down. One gate is enforced at the prompt — sexual content involving minors — and it cannot be disabled. Everything above that floor is your policy to set.

Are generated images filtered before I receive them?

No. There is no post-generation classifier here, so a job never fails for what it produced and you are never billed for an image you are not shown.

How hard is the migration?

The main change is shape rather than schema: the images endpoint there is synchronous, and here you submit a job and collect the result by poll or webhook. In most codebases that is one client module plus a callback route.

FastGen Cloud vs the OpenAI image API

The OpenAI image API is a strong general-purpose generator with a policy layer built into the product: the model revises your prompt before rendering it, prompts and outputs are filtered, and C2PA provenance metadata is embedded in what comes back. FastGen Cloud makes the opposite trade — your prompt reaches the sampler as written, and the policy above the legal floor is yours to set.

Get an API key Quickstart

Side by side

Dimension	FastGen Cloud	OpenAI
Prompt handling	Verbatim. The string you send is the string the sampler conditions on — no rewriting and no expansion.	The mainline model automatically revises your prompt before generation; the rewritten text is surfaced back to you as revised_prompt.
Content filtering	A single minor-safety gate on the prompt. Adult, violent and political content passes; output is never classified.	Prompts and generated images are filtered under the platform content policy. A moderation parameter accepts auto (default) or low — there is no off.
Provenance metadata	None embedded. If your product needs a provenance claim you add it downstream, where you control it.	C2PA metadata is included in generated images.
Model catalogue	A curated set of open-weight, commercially licensed checkpoints, each behind a stable model id.	OpenAI’s own image models.
Identity from a photo	A dedicated Face ID endpoint from one reference photo, plus character training from 4–30 photos referenced by id.	No dedicated identity endpoint; likeness is steered through prompting and image input.
Job lifecycle	202 with a job id, then poll or receive a signed webhook. Idempotency-Key replays rather than double-charging.	Synchronous request/response — you hold the connection until the image is returned.
Payment model	Prepaid balance, debited at submit, auto-refunded when a job fails. A bug cannot run up a bill.	Postpaid usage billed against your account.
Video	Image-to-video in the same API, billed per second of output.	Offered through separate video models and endpoints.

Comparison of platform design, not of prices — OpenAI sets its own rates and changes them independently, so check its current pricing page before deciding. Last reviewed 22 September 2026.

When the OpenAI image API is the better choice

Worth saying plainly: for a large class of products it is, and swapping providers to avoid a prompt rewrite you do not mind is not a good reason to migrate.

You want strong instruction-following from plain conversational English, without prompt-engineering the way diffusion checkpoints expect
You are already on the OpenAI SDK and want one vendor, one key and one invoice
Embedded C2PA provenance is something you want rather than something you work around
Your content sits comfortably inside a mainstream content policy, so the filter never fires anyway

When verbatim prompts matter

Prompt revision is a quality feature for a single interactive user and a correctness problem inside a product. If the platform rewrites non-deterministically, a fixed seed stops fixing the image, prompt A/B tests measure the wrong thing, and a bad result could be your prompt or the rewrite with no way to tell which.

You need reproducibility — same prompt and seed, same image, every time
You are testing prompt wording and need the thing you ship to be the thing you tested
You are building white-label output and do not want injected house style or rating tags
Your product serves adults and a mainstream content policy is the binding constraint rather than model quality

What migrating actually involves

The shape is the same everywhere: authenticate, submit a job, wait for a callback, download the output. In practice a migration is one client module and a model-name mapping.

submit + wait

const res = await fetch("https://api.fastgencloud.com/v1/images/generations", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.FASTGEN_API_KEY}`,
    "Content-Type": "application/json",
    "Idempotency-Key": requestId,
  },
  body: JSON.stringify({ model: "general-image", prompt, size: "1024x1024" }),
});
const job = await res.json();          // 202 → { id, status: "queued" }

// then either poll GET /v1/jobs/{id}, or pass webhook_url and get a signed callback

How billing works here

Money is a prepaid balance in your account. You top it up with a card, each job debits it at submit time, and a failed job is refunded automatically — there is no invoice at the end of the month and no way to run up a bill you did not intend.

Rates are per image or per second of video, published on the pricing page. Images above 1 megapixel bill pro rata by pixel count. Outputs are retained for up to 7 days.

Frequently asked questions

Does FastGen Cloud rewrite my prompt like revised_prompt does?: No. There is no language model in front of the sampler. The only strings prepended are a checkpoint’s native quality tags and a LoRA trigger word when you select a catalog LoRA, both documented on the model page.
Is there a moderation setting I can turn down?: There is no dial, because there is no general content filter to turn down. One gate is enforced at the prompt — sexual content involving minors — and it cannot be disabled. Everything above that floor is your policy to set.
Are generated images filtered before I receive them?: No. There is no post-generation classifier here, so a job never fails for what it produced and you are never billed for an image you are not shown.
How hard is the migration?: The main change is shape rather than schema: the images endpoint there is synchronous, and here you submit a job and collect the result by poll or webhook. In most codebases that is one client module plus a callback route.

Start generating

Create an account, add a prepaid balance, and call the API with a key from the console. No subscription, no minimum, no per-seat pricing.

Get an API key