API · Private beta

Six AI image models.
One endpoint.

Programmatic access to Nano Banana 2, Flux, Ideogram, Recraft, Imagen and DALL-E 3. Switch models per request, keep one brand kit across all of them, and see per-image cost on every response. From $100/mo for 5,000 calls.

Join the waitlistSee pricing
One endpoint, six modelsBrand kit persistenceTransparent per-image cost$50 in free credits
The request

One endpoint, six models

Switch models by changing one parameter. Same auth, same brand kit, same response shape — no per-vendor SDK, no rewrite.

POST /v1/images/generate
curl https://api.clearly.sh/v1/images/generate \
  -H "Authorization: Bearer <YOUR_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Linen pillow in cream, mid-century lounge chair, golden hour",
    "model": "flux-1.1-pro",
    "aspectRatio": "4:3",
    "brand_kit_id": "kit_abc123"
  }'

# Response:
{
  "id": "gen_xyz789",
  "url": "https://files.clearly.sh/u/abc/gen_xyz789.png",
  "model": "flux-1.1-pro",
  "cost_usd": 0.04,
  "prompt_used": "[brand-kit context] + [prompt]"
}
The model menu

Pick the right model per generation

Best-in-class for every job — photoreal product shots, text-in-image, vector art, illustration — all priced per image, all behind the same key.

FeatureModel IDBest forPer image
Nano Banana 2nano-banana-2Fast, cheap, high-volume$0.039
Flux 1.1 Proflux-1.1-proPhotoreal product shots$0.040
Ideogram v2ideogram-v2Text-in-image & posters$0.050
Recraft v3recraft-v3Vector & brand-consistent art$0.060
Imagen 4imagen-4Google-grade realism$0.070
DALL-E 3dalle-3Creative & illustrative$0.080

Retail per-image rates. Bundled into your monthly tier; overage rates apply past your cap.

Why this API

Built for products, not playgrounds

One integration that keeps your brand consistent, your costs auditable, and your model choice open.

Brand kit, every model

Create a brand kit once — color, font, reference photos, style — and attach it by id to any generation. The same identity carries across all six models, no re-prompting per vendor.

Transparent per-image cost

Every response returns cost_usd. Audit spend by user, brand, or product in the dashboard — no end-of-month surprise from an opaque credit system.

Async webhooks

Pass callback_url for long-running jobs. Flux Pro HD can take 30-60s; Nano Banana 2 is typically under 5s — either way, we POST the result back.

Resell to end users

Pass a user_id per request to track per-user usage and bill your own customers. Ship image gen inside your app, on our infrastructure.

How it works

Key, call, ship

Step
01

Get a key

Join the waitlist. Approved teams get an API key and a Stripe-billed account within 1-3 business days — first 100 get $50 in free credits.

Step
02

Make one call

POST a prompt and a model to /v1/images/generate. Switch models with a single parameter; attach a brand_kit_id to stay on-brand.

Step
03

Ship & audit

Get a URL and the exact cost back per request. Track usage per user, brand, or product — and resell generations inside your own app.

Pricing

Per-call pricing that scales down per image

Monthly tiers bundle your calls; overage rates fall the more you use. Enterprise is per-pixel-priced with an SLA.

FeatureDeveloperProductionScaleEnterprise
Price$100/mo$500/mo$2,000/moCustom
Included5,000 calls / mo50,000 calls / mo500,000 calls / moCustom
Overage$0.025 / call$0.015 / call$0.008 / callNegotiated
Best forIndie devs + side projectsSaaS products + agenciesHigh-volume platformsPer-pixel-priced workloads + SLA
99.5% SLA on Production & ScaleBurst tolerance 2x for 10sPer-request billing audit
FAQ

API, answered

01When can I get API access?+
Public API is in private beta — joining the waitlist puts you in front of the queue. Approved teams get an API key + Stripe-billed account within 1-3 business days. Currently ~50 teams on the waitlist; first 100 get $50 in free credits.
02What does the request look like?+
A POST to /v1/images/generate with { prompt, model, aspectRatio?, seed?, brand_kit_id? }. Response is JSON with { url, model, cost_usd, prompt_used }. The same endpoint serves all six models — change the `model` parameter to switch (e.g. "flux-1.1-pro", "ideogram-v2", "recraft-v3", "imagen-4", "dalle-3", "nano-banana-2").
03How is per-image cost calculated?+
Per-model rates ($0.039-$0.080 per image at retail) are bundled into the tier monthly. Once you hit your monthly cap, overage rates kick in (see table above). Detailed billing is shown per-request in the dashboard so you can audit costs by user, brand, or product.
04Do you offer a brand kit API?+
Yes — create brand kits via POST /v1/brand-kits, attach them to generations via `brand_kit_id`. The brand kit carries color, font, reference photos, and style preferences. Re-usable across generations and across models.
05What about webhooks for long-running generations?+
Yes — async mode. Pass `callback_url` in the request body and we POST the result back when generation completes. For Flux Pro at HD resolution this can take 30-60s; for Nano Banana 2 typically <5s.
06Do you offer SLA?+
Production and Scale tiers include 99.5% uptime SLA. Enterprise can negotiate 99.9% with dedicated infrastructure. SLA credits apply against your monthly bill if missed.
07How is rate limiting handled?+
Soft cap of 60 requests/min on Developer, 600/min on Production, 6,000/min on Scale. Bursts up to 2x for 10 seconds are tolerated. Enterprise gets custom rate limits.
08Can I use this for end-user-facing image gen in my app?+
Yes. You can resell generations to your end users as long as you stay within Clearly's acceptable use policy (no unauthorized commercial likeness, no CSAM, no copyright infringement at scale). Pass a `user_id` per request so we can track per-user usage for your billing.
Explore the platform

Built on the Clearly stack

Company Brain

The knowledge layer your agents recall and execute — vector memory, private by default.

Beehaven CLI

The terminal + MCP interface to your agent cloud — drive generations from the command line.

Vector Studio

Turn prompts into editable vector art on a canvas — the visual side of the same engine.

Build on six models, one key

Join the private-beta waitlist for an API key and Stripe-billed account. First 100 teams get $50 in free credits.

Join the waitlistSee web pricing ⟶