Antimatter

POST/v1/images/generationsAPI key

Served on the terms a completion is: the key's chat scope and policy, the account's credits with a hold at the worst case, and one metered usage row settled at what the upstream measured. A model with no published price, or one priced in a unit the endpoint does not report, is refused on house money with 503 pricing_unavailable and served on your own provider key. With no provider configured the answer is 503 upstream_unavailable. No fallback chain and no x402: a media request names one model and needs a key. An image from OpenRouter is charged the cost its answer reports; an OpenAI image is charged per token of each kind at OpenAI's list price. The hold reserves 8192 output tokens per image. A caller who disconnects does not cancel a generation that began: it is finished and metered.

Request body

application/json, required

Request body fields
Field	Type	What it is
modelrequired	string	An image model: openai/gpt-image-* on the native OpenAI adapter, or any model OpenRouter's /images endpoint lists. GET /v1/models marks them with output_modalities image. At most 200 characters.
promptrequired	string	At most 32000 characters.
n	integer	From 1 to 10.
size	string
quality	string	One of "auto", "low", "medium", "high", "standard", "hd".
background	string	One of "auto", "transparent", "opaque".
response_format	string	b64_json (default) or url. The gateway keeps no blob storage, so url answers a data: URL holding the image itself. One of "b64_json", "url".
output_format	string	One of "png", "jpeg", "webp".
output_compression	integer	From 0 to 100.
moderation	string	One of "auto", "low".
user	string	At most 256 characters.
aspect_ratio	string	OpenRouter only; named in x-antimatter-adjusted when the model is served elsewhere.
resolution	string	OpenRouter only, as aspect_ratio. One of "512", "1K", "2K", "4K".
stream	boolean	Refused when true: streamed generation is not offered.
partial_images	integer	Refused: streamed generation is not offered. From -9007199254740991 to 9007199254740991.

Example

{
  "model": "antimatter/auto",
  "prompt": "Say hello in five words."
}

Answer 200

The images, and the token counts where the upstream reported them.

Headers: x-antimatter-upstream, x-antimatter-model, x-antimatter-request-id, x-antimatter-byok, x-antimatter-adjusted, x-ratelimit-limit, x-ratelimit-remaining, x-ratelimit-reset

Fields of the 200 answer
Field	Type	What it is
created	integer
data	array of object
data[].b64_json	string
data[].url	string	A data: URL.
data[].revised_prompt	string
usage	object
usage.input_tokens	integer
usage.output_tokens	integer
usage.total_tokens	integer

Errors

400OpenAI shaped error.
401OpenAI shaped error.
402OpenAI shaped error.
403OpenAI shaped error.
404OpenAI shaped error.
413OpenAI shaped error.
429OpenAI shaped error.
502OpenAI shaped error.
503OpenAI shaped error.
504OpenAI shaped error.

Every error answers in one shape, with its type, code and the field it names. The whole list is in the docs' Errors section.

Try it

The same call with curl

curl -X POST https://useantimatter.com/v1/images/generations \
  -H "Authorization: Bearer $ANTIMATTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"antimatter/auto","prompt":"Say hello in five words."}'