Every AI model.
Twelve bills.
One key.
Chat, images, video, voice, music and transcription through a single API key and a single prepaid balance, priced below the vendors’ own list prices.
Pay as you go from $5. No subscription. Every call shows exactly what it took from your credit.
# one key, one base URL: https://omni-api.rcinc.app/v1
One API in front of the AI models you already use.
The Omni API is a single API and billing account for many AI models. Instead of opening accounts, keys and invoices with OpenAI, Anthropic, Google, ElevenLabs, video model vendors and others, you create one Omni key, add credit, and call any of them through the same base URL.
It is built for developers and product teams that use several AI models. It accepts the OpenAI and Anthropic request formats, runs image, video and audio work as tasks with webhooks, and reports the cost of every call. It is made by Radiant Crest Inc.
One API.
Six kinds of output.
Hundreds of models from OpenAI, Anthropic, Google, ByteDance, Kling, MiniMax, ElevenLabs, Suno and more. New ones are added as they launch, and your key already works with them. See all features.
Every frontier model, OpenAI-compatible
GPT, Claude, Gemini, DeepSeek, Grok and more behind /chat/completions, /responses and /messages, with streaming. Point any OpenAI or Anthropic SDK at Omni and change nothing else.
POST /v1/chat/completionsImagesGPT Image 2, Nano Banana, Seedream
Text-to-image and edits with up to ten reference images.
POST /v1/images/generationsVideoVeo, Kling, Seedance, Hailuo, Wan
Text and image to video, lip-sync and upscaling, as async tasks with webhooks.
POST /v1/videos/generationsVoiceThousands of voices, plus your own
Text to speech, multi-speaker dialogue, voice cloning, voice changer and dubbing.
POST /v1/audio/speechMusic & soundSongs, beds and effects
Full songs with lyrics or instrumentals, and sound effects up to 30 seconds.
POST /v1/audio/musicTranscriptionSpeech to text with word timings
JSON and SRT out, with audio-event tagging, for files up to 200 MB.
POST /v1/audio/transcriptionsSame models.
Less than going direct.
Compare what you pay on Omni with each vendor’s own public list price for the same model. You pay per call from one balance; there is nothing to subscribe to. Full pricing.
average saving on frontier language models against OpenAI, Anthropic and Google list prices.
on GPT Image 2 (per 1K image) compared with OpenAI.
GPT-6.1 Sol output tokens for $20, the price of one ChatGPT Plus month, with no monthly fee.
| Model | Going direct (list price) | On Omni | You save |
|---|---|---|---|
| Language models | |||
| GPT-6.1 Solper 1M tokens, in / out | $1.49 / $7.47 | −25% | |
| Claude Opus 5.5per 1M tokens, in / out | $2.99 / $14.95 | −25% | |
| Claude Sonnet 5per 1M tokens, in / out | $1.51 / $7.55 | −25% | |
| Gemini 3.8 Flashper 1M tokens, in / out | $0.561 / $2.80 | −25% | |
| Images | |||
| GPT Image 2per 1K image | ≈ $0.0067 | −97% | |
| Nano Bananaper image | $0.023 | −41% | |
| Seedream 5 Proper 2K image | $0.081 | −11% | |
| Video | |||
| Veo 3.1 Fastper 1080p clip | $0.374 | −69% | |
| Veo 3.1 Qualityper 1080p clip | $1.47 | −54% | |
| Kling 3.0 Omniper second, 720p | $0.081 | −4% | |
| Hailuo 2.3 Standardper 6 s clip, 768p | $0.172 | −38% | |
| Voice & sound | |||
| ElevenLabs Multilingual v2per 1K characters | $0.069 | −14% | |
| Sound effect, 5 secondsper effect | $0.0007 | −93% | |
| AI song (Suno)per song | ≈ $0.0049 | −75% | |
Vendor prices are their public list prices, checked 15 Oct 2026; the Suno figure is its Pro plan ($10 for about 500 songs). Omni prices are what you are charged today. Image, video and audio calls are billed on the exact work each job does, so prices marked ≈ are typical and a job can cost less. Taxes not included.
Live in three minutes.
Sign in with Google or your email, add credit, create a key. Existing OpenAI and Anthropic code works by changing two lines. Step-by-step guide.
Sign in
Google sign-in, or a six-digit code sent to your email. No passwords to manage.
Add credit
Top up any amount from $5 by card. One balance pays for every model; it never expires.
Call any model
Create a key, set the base URL, and watch each call’s deduction in your dashboard.
from openai import OpenAI
client = OpenAI(api_key="omni_live_…", base_url="https://omni-api.rcinc.app/v1")
reply = client.chat.completions.create(
model="gpt-6.1-sol",
messages=[{"role": "user", "content": "Summarise this changelog"}],
)
print(reply.choices[0].message.content)
print(reply.omni["charged_usd"]) # what this call took from your creditWhen Omni is the right choice.
An honest view of where one aggregated API helps and where going direct may suit you better. Full comparison.
Choose Omni when you
- use models from more than one vendor and want one key, one balance and one usage report;
- want lower prices than list without committing to a plan;
- need image, video, voice and music generation alongside text.
Go direct when you
- need a vendor’s newest beta features on day one, or a direct contract and support agreement with that vendor;
- need guaranteed capacity or data-residency commitments that only the vendor offers;
- use a single model at very high volume with your own negotiated rate.
Frequently asked questions
What is the Omni API?
The Omni API is a single API for many AI models. One key and one prepaid balance give you language models (GPT, Claude, Gemini and others), image and video generation, text to speech, voice cloning, music, sound effects and transcription. It is built by Radiant Crest Inc.
How much does it cost?
There is no subscription. You add credit from $5 by card, and each API call deducts its price. Language models cost about 25% less than the vendors' own list prices, and image, video and audio jobs are often far cheaper. The pricing page lists current prices.
Is it compatible with the OpenAI SDK?
Yes. The API accepts the OpenAI Chat Completions and Responses formats and the Anthropic Messages format, so existing OpenAI and Anthropic SDKs work by changing the base URL to https://omni-api.rcinc.app/v1 and using your Omni key.
How do I know what each call cost?
Every text response includes omni.charged_usd and an x-omni-charged-usd header, every image, video or audio task shows its charge, and the dashboard lists each deduction.
Do you store my prompts or outputs?
No. Prompts, files and generated outputs pass through to the model provider and back. Omni keeps only the result links shown on a task (removed after 90 days) and billing records, and never uses your content to train models.
What are the limits?
Each key can make 600 requests a minute by default. Image, video and audio jobs run as tasks that take from seconds to several minutes. Direct file uploads are limited to about 4 MB; larger files are sent as a public link.
Do unused credits expire?
No. Credit stays in your balance until you use it, and unused credit is refundable on request within 30 days of purchase.
Stop juggling
AI accounts.
One key, one balance, one usage report, and lower prices on the models you already use.
Get your API key →