Flux API key: two paths, and what changes after you have one
·
Flux Pro v1.1 runs upstream on fal. You can hold a key with the vendor directly, or one key that reaches Flux Pro v1.1 alongside the other models through a single interface. The tradeoffs are laid out in how to get an AI video API key — this page covers what changes once you have one.
What a call costs
5 credits at default settings, which is $0.05–$0.07 depending on your pack.
The constraint to know before you build
n caps at 4. Cost is linear, so batching buys concurrency slots rather than a discount.
Using it
export GENACE_API_KEY=sk_xxx
curl -X POST https://genace.ai/api/v1/images/generations \
-H "Authorization: Bearer $GENACE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"flux-pro","prompt":"..."}'
Verifying the key costs nothing
curl https://genace.ai/api/v1/models -H "Authorization: Bearer $GENACE_API_KEY"
This authenticated read has no generation charge and skips generation concurrency checks, but its own request-rate limit still applies. A 200 means the key is live; anything else is covered in MISSING_API_KEY and INVALID_API_KEY.
Understand the quota scopes
| Tier | Requests per minute | Concurrent jobs |
|---|---|---|
| Free | 5 | 1 |
| Pro | 60 | 5 |
Request rate is enforced per API key; concurrent generation jobs are counted per user. A second key does not create another user concurrency allowance. Separate keys can still isolate credentials and key-specific spending caps.
Continue with models and parameters
Models covered on this page
Where these facts come from
- codebase: src/ai/api/wrap.ts
- codebase: src/ai/providers/flux.ts — pricing() and paramsSchema
- codebase: src/ai/api/auth.ts — key verification
- codebase: src/ai/api/quota.ts — per-tier ceilings