Pay as you go
Per token for text, per image for images. One prepaid balance and one set of API keys for both. No subscription, no minimum commitment.
Text
DeepSeek V4.1 Flash
deepseek-ai/DeepSeek-V4.1-Flash| Tokens | USD per million |
|---|---|
| Input | $0.1425 |
| Input, cache hitrepeated prompt prefix | $0.0029 |
| Outputreasoning tokens included | $0.5700 |
- One price at every hour of the day.
- A repeated prompt prefix (the system prompt, earlier turns) is billed at the cache-hit rate; usage.prompt_tokens_details.cached_tokens in each response says how many tokens hit.
- The model reasons before it answers; reasoning tokens are billed as output.
GLM-5.3-Flash
Pricing will be published here when the model goes live.
Images
Ideogram 4.5
ideogram-4-5images.luminal.cloud| Quality | USD per image |
|---|---|
very_lowsource images required | $0.0080 |
low | $0.0300 |
mediumdefault for edits | $0.0600 |
highdefault for text-to-image | $0.2000 |
- These are the same prices Ideogram charges for its own API. Luminal adds no markup.
- Billed per returned image, the same at every size.
- quality picks the tier. Text-to-image with no quality set is billed as high; an edit (Generate with source images, or Precise edit) with no quality set is billed as medium.
- num_images counts the outputs you receive: four images at medium cost 4 × $0.0600 = $0.2400. Best-of tiers do not bill their internal candidates.
- Images withheld by the provider's safety or copyright detection (is_image_safe=false) are not billed.
- When you submit, the full request is held against your balance; when it completes, you are charged for the images actually delivered and the rest of the hold is released. Credits left dips by the hold while generations are in flight.
- A synchronous request that runs longer than 10 minutes returns 202 with a generation_id instead of the images — poll GET /v1/generations/{generation_id} on the same host, with your key, until it completes.
Same endpoints, fields and file rules as Ideogram's API, under https://images.luminal.cloud, with your Luminal key as Api-Key or a bearer token. What differs from calling Ideogram directly is documented at https://images.luminal.cloud/docs/.
How billing works
- Add credit by card from the dashboard's billing page; the minimum top-up is $50.00.
- Each completed text response is charged when it finishes, from the token counts in its usage field.
- Each image generation is held when it is submitted and charged for the images delivered when it completes.
- Your balance and day-by-day usage are on the dashboard and at GET /v1/balance on either host.
- Credit does not expire.
Default limits
TEXT
| Requests per minute | 60 |
| Tokens per minute | 1,000,000 |
| Burst | 10 requests per second |
IMAGES
| Requests per minute | 60 |
| Images per minute | 120 |
| Generations in flight | 8 |
| Status polls per minute | 600 |
Need more? Write to founders@luminal.com with your expected load.