Don't speak API? Just describe what you want.
Every image, video, and voice model on our gateway is available inside the browser. Type a prompt, click Generate, download the result. No SDKs, no setup, no terminal.
Every media model — one endpoint.
Hover any tile to see the top models we recommend.Tap any tile to see the top models for each capability. Swap the model string — keep your code, prompts, tools.
Image generation
Concept art, product mocks, brand assets. One endpoint, every frontier model.
Vector & SVG
Text-to-vector graphics — logos, icons, illustrations. Scalable, editable, ready for production.
Image editing & upscale
Image-to-image edits, outpainting, AI upscalers. Up to 8× detail recovery on stills.
Video generation
Generate, extend, image-to-video. Frontier quality from every major lab.
Music & song
Generate full music tracks from text prompts — vocals, instruments, up to 4 minutes. 1-4 variants per request.
Speech → text
Diarization, timestamps, 99+ languages. Realtime or batch.
Text → speech
10 ElevenLabs TTS models + Deepgram Aura 2. 30+ languages, 100s of voices per model.
Realtime voice
Sub-300ms voice agents. Two-way audio with function calling.
Every leading media model.
Zero contracts.
Image, video, voice, realtime — switch by changing one string.
You pay $50.
You get $100 of compute.
Our buying power becomes your pricing edge. Then every request bills at the upstream's actual cost — no markup, no padded units.
How $50 becomes $100.
You buy $100 of credit
Pay $50 in crypto or via redeem code. We seed your wallet with $100 of upstream credit immediately.
Every job bills at upstream cost
An image generation that costs us $0.04 bills your wallet for $0.04. Itemized in USD, per image / per second / per minute — no rounding tricks.
No expiry, no minimums after
Credit doesn't disappear. Re-top whenever — same 50%-off ratio applies.
launch pricing · while supplies last ·first 1,000 wallets
Three lines.
That's the migration.
Change the base URL. Change the API key. Swap the model name to any media model in the catalog. Your code, your prompts, your tools — unchanged.
- 100% schema-compatible with the OpenAI SDK
- Image, video, audio — same endpoint shape
- Per-job webhooks for long-running renders
- Signed asset URLs · automatic retries
from openai import OpenAI
client = OpenAI(
base_url="https://api.echotokens.com/v1",
api_key="sk-echo-...",
)
response = client.images.generate(
model="flux-2-pro",
prompt="a serene mountain lake at dawn",
n=1,
size="1024x1024",
)
print(response.data[0].url)Engineered for production.
Priced like a deal.
Fallback. Webhooks. Signed URLs. Async queue. Spend limits. IP locks. The infrastructure you'd build yourself — already built.
Smart fallback
Upstream rate-limited? We retry on a sibling model. You set the priority list.
Per-job webhooks
Long-running video and image jobs ping your endpoint when ready. No polling, no timeouts.
Signed asset URLs
Generated images and videos return signed URLs scoped to your project. Expire in 24h or shorter on demand.
Asynchronous job queue
Video renders can take minutes. Submit, poll, or webhook — pick your retrieval style per job.
Spending limits
Cap a key at $5/day or $50/month. Hard ceilings, soft warnings, rolling windows.
Itemized usage
Every job: model, asset count, seconds rendered, USD cost, latency, request ID — exportable as CSV.
IP allowlisting
Lock keys to specific CIDR ranges. Per-key, per-environment, zero downtime to update.
Bring your own storage
Pipe generated assets straight to your S3 / GCS bucket. The gateway never holds your media.
Ship in five minutes.
Sign up, drop in a key, point your SDK. The next hundred dollars of compute is on us — at fifty.