One API for every AI model. Pay per generation.
162 image, video, audio and 3D models behind one key — Veo 3.1, Seedance 2.5, Kling 4.0, GPT Image 2.5, Nano Banana Pro, ElevenLabs and more. Priced per request, no minimum, no subscription, and the exact cost quoted before every call.
- 162 models, one key
- Price quoted before every call
- Failed generations refunded
- Credits never expire
Built for teams that ship
Every model, one integration
One endpoint, one key and one balance across 162 models from Google, OpenAI, ByteDance, Kuaishou, Alibaba, MiniMax, Runway, ElevenLabs and more. New models appear on the same key within days of release.
Know the cost before you run
Every request can be priced first: send the same body to the public price endpoint and get the exact credits back before you spend anything — no account needed to ask.
Pay per generation, nothing else
No subscription, no seat fee, no minimum commitment. Buy credits once by card or crypto; they never expire, a failed generation is refunded automatically, and each API key can carry its own monthly spending cap.
SDKs, webhooks, MCP
JavaScript and PHP SDKs, an OpenAPI spec and Postman collection, completion webhooks instead of polling, and an MCP connector so your team can call the same models from Claude, ChatGPT or Cursor.
Throughput from day one
Up to 10 generations in flight per account and 300 API requests a minute out of the box — no plan upgrade to unlock concurrency.
Hosted outputs and history
Every output is delivered on a CDN URL you can hand straight to your product, with the full request history kept on the account. Generated files stay available for 30 days, so copy what you keep.
API prices by model
What one request costs on Flixly, in USD. The Flixly figures are quoted live by the same engine that bills the API, so they cannot drift from the charge.
| Model | What is priced | Credits | USD at list | USD at volume |
|---|---|---|---|---|
| Video | ||||
| Veo 3.1Google | 8-second video, 1080p, with audio | 534 | $5.34 | $4.81 |
| Veo 3.1 FastGoogle | 8-second video, 1080p, with audio | 200 | $2.00 | $1.80 |
| Seedance 2.5ByteDance | 5-second video, 1080p | 970 | $9.70 | $8.74 |
| Seedance 2.0ByteDance | 5-second video, 1080p | 569 | $5.69 | $5.13 |
| Kling 4.0Kuaishou | 5-second video | 178 | $1.78 | $1.60 |
| Kling 3.0Kuaishou | 5-second video | 85 | $0.85 | $0.77 |
| Kling 3.0 TurboKuaishou | 5-second video | 71 | $0.71 | $0.64 |
| Wan 3.0Alibaba | 5-second video, 1080p | 117 | $1.17 | $1.05 |
| Runway Gen 4.5Runway | 5-second video | 63 | $0.63 | $0.57 |
| MiniMax H3 MaxMiniMax | 6-second video, 1080p | 160 | $1.60 | $1.44 |
| Image | ||||
| GPT Image 2.5 FlareOpenAI | One image | 17 | $0.17 | $0.15 |
| Nano Banana ProGoogle | One image | 25 | $0.25 | $0.23 |
| Nano Banana 2Google | One image | 14 | $0.14 | $0.13 |
| Seedream 5.0 ProByteDance | One image | 13 | $0.13 | $0.12 |
| Ideogram 4.5Ideogram | One image | 10 | $0.10 | $0.090 |
| FLUX 2 FlexBlack Forest Labs | One image | 12 | $0.12 | $0.11 |
| Qwen Image 2.1Alibaba | One image | 4 | $0.040 | $0.036 |
| Audio | ||||
| ElevenLabs Multilingual v2ElevenLabs | 1,000 characters of speech | 17 | $0.17 | $0.15 |
| Gemini 3.8 Flash TTSGoogle | 1,000 characters of speech | 8 | $0.080 | $0.072 |
| 3D | ||||
| Meshy 7.1 Text-to-3DMeshy | One 3D model | 200 | $2.00 | $1.80 |
Prices in USD, updated hourly. Longer, higher-resolution or multi-image requests cost more; quote any request exactly with the price endpoint before running it. Full catalogue on the models page.
Three lines to your first generation
Create a key in the dashboard, send a request, and read the output URL. Every model takes the same call.
curl https://www.flixly.ai/api/v1/generate \
-H "Authorization: Bearer flx_live_YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "nano-banana-pro",
"type": "TEXT_TO_IMAGE",
"prompt": "Product shot of a ceramic mug on oak, soft daylight"
}'Who builds on Flixly
- Products that generate images or video for their own users — e-commerce visuals, avatars, ads, thumbnails — without running a model fleet.
- Agencies and studios that need every frontier model in one place, with one invoice and a price shown before each run.
- Teams working in Claude, ChatGPT or Cursor who want the same models and the same balance inside their tools via MCP.
- Developers who want to pay for what they generate, with no monthly plan to justify in a quiet month.
Questions teams ask
Is there a minimum spend or a subscription?
No. You buy credits once, spend them per generation, and they never expire. There is no seat fee, no monthly plan and no minimum commitment.
How do I know what a request will cost before I send it?
Send the same body to the public price endpoint. It returns the exact credits that request will cost, so you can show the price to your own users or budget a batch before running it.
What happens when a generation fails?
The credits are returned to your balance automatically. You are only charged for outputs you receive.
What are the rate and concurrency limits?
Every account gets up to 10 generations in flight at once and 300 API requests a minute. If you need more, contact us with your expected volume.
Can my company get invoices?
Every card purchase comes with an itemised receipt for the pack. For invoices, purchase orders or custom terms at volume, contact us.
Can we use the outputs commercially?
Yes. You own what you generate on Flixly, including for ads, products and client work, subject to the terms of each model's creator.
What happens to our prompts and uploads?
They are used to run your generations and kept in your account history so you can reuse them. Flixly does not sell your content. Details are in the privacy policy.