Your AI operator · solution partner
Not just an API —
Not just an API —
your AI solution partner
We're not a 'wrapper'. With OUR OWN models trained for your industry, ready AI tools and an end-to-end platform, we deliver solutions that work for your business — not just an API key. Models, chat, RAG, automation and memory in one panel — pay as you go or subscribe; or on your own servers (on-prem).
No card required · OpenAI-compatible · your first request in minutes
Access with a single API key
Write a short product description: wireless earbuds.
gpt-4o · 0.4 s
Crystal-clear sound, up to 30 hours of battery and active noise cancellation — with you all day.
Crystal-clear sound, up to 30 hours of battery and active noise cancellation — with you all day.
20+
models available
1+
AI providers
%99.9
target uptime · automatic failover
One platform
gateway · tools · memory · on-prem
Not a gateway, a platform
Much more than a model gateway
Other platforms connect you to models and stop there. Operayn adds, on top of the models, ready tools your team uses today, memory and enterprise deployment — that's our difference.
| Model gateway only | The Operayn platform | |
|---|---|---|
| One API for all models | ||
| Automatic failover & smart routing | ||
| Ready chat interface (files/vision) | ||
| Knowledge bases (RAG) with citations | ||
| Custom assistants & system prompts | ||
| Visual automations (AI Studio, 570+ templates) | ||
| Three-layer AI memory | ||
| Custom-trained / expert models | ||
| Team, roles & 30+ language UI | ||
| On-prem / air-gapped · compliance & e-invoicing |
One API, one invoice
Pick the model you need, we handle the rest
No more separate accounts, payments and integrations across providers. Operayn unifies them behind a single OpenAI-compatible API; with your single API key you reach them all — you just change the model name.
Anthropic
Access to 20 models with one key · the list keeps growing
One line to switch models:
from openai import OpenAI
client = OpenAI(base_url="https://<alan-adınız>/v1", api_key="opn-...")
client.chat.completions.create(model="gpt-4o", messages=[...]) # ya da
client.chat.completions.create(model="claude", messages=[...]) # ya da
client.chat.completions.create(model="ozel-model", messages=[...]) # size özelModel catalog (20+)
claude-fable-5.1[1m]
claude-fable-5.1:batch[1m]
deepseek/deepseek-v4.1-flash[1m]
google/gemini-3.8-flash[1m]
google/gemini-3.8-flash:batch[1m]
ibm-granite/granite-4.2-8b
inception/mercury-2.5
inclusionai/ling-3.0-flash-fin
inclusionai/ling-3.0-flash-sante:free
inclusionai/ling-3.0-flash-vl:free
meta/muse-spark-1.3[1m]
meta/muse-spark-1.3-contributor[1m]
nex-agi/nex-n2.5-mini:free
nex-agi/nex-n2.5-pro:free
openai/gpt-6-astra[1m]
openai/gpt-6-astra:batch[1m]
openai/gpt-6-astra-pro[1m]
openai/gpt-6-astra-pro:batch[1m]
qwen/qwen3.8-max-0902
tencent/hy4-preview[1m]
No matching model.
The list keeps growing — reach them all with one key by just changing the model name. Plus, we also train models specifically for you.
Features
Why Operayn?
All models in one place
GPT, Claude, Gemini, Llama and more — one subscription, one invoice, one integration.
No downtime
If a model or provider fails, it automatically fails over to a backup. Your service keeps running.
Transparent & flexible billing
Pay as you go. Wallet, quota and subscription options; no surprise bills.
Models we train for you
We're not just a middleman — we deliver expert models WE train for your industry (legal, e-commerce, health and more), from the same platform.
Batch processing
Process many requests at once, asynchronously and at a discount.
Integrate in 5 minutes
OpenAI-compatible. Just change base_url in your existing code — it works instantly.
Speed & caching
Smart caching returns faster, cheaper responses for repeated requests.
Security & compliance
Encrypted secrets, isolated database, 2FA and audit logs. Your data stays safe.
Usage analytics
Track your spend, requests and model breakdown transparently from the dashboard.
Data control & sovereignty
You decide which provider receives which data. PII masking, KVKK/GDPR and an isolated database; with on-prem, data never leaves your organization.
Ready-made AI tools
Put AI to work without writing code
Not just an API; a ready AI platform your team can use today. All with your own data, your own brand, from one panel.
Chat interface
A ready chat screen with no setup. Upload images and files (PDF, Word, Excel), keep history — no technical knowledge needed.
Knowledge bases (RAG)
Upload your own documents; let AI answer based only on your data, citing sources.
Your own assistants
Derive your own custom assistant from platform models: your own instructions, your own knowledge base, your own brand.
Visual automations (AI Studio)
Build agentic workflows on a drag-and-drop canvas. 570+ ready business templates — pick one, adapt it, run it.
Team & roles
Invite your team, assign roles and permissions. Scale your user (seat) count as you need.
Interface in 30+ languages
Your team uses the platform in their own language. Interface and help content in 30+ languages.
AI memory
Your AI now remembers — three-layer memory
No conversation has to start from scratch. Operayn offers three separate memory layers: model-specific memory, personal (seat) memory, and company-wide corporate memory. All are kept in isolated, encrypted vector stores; who writes and who reads each layer is governed by CLEAR permission rules.
Model memory
Enable it on a model you derive; every conversation to that model (chat or API) is stored, and relevant memories are recalled automatically on later requests. Can be isolated per user.
Personal (seat) memory
A personal, cross-model AI memory for every employee. Whatever model they use, preferences, projects and working context are remembered — no re-explaining every time, so everyone works faster. Personal data is encrypted, isolated and privacy-compliant; control and deletion always stay with the user.
Corporate memory Enterprise
The company's SHARED brain. Knowledge across all your models, chats and RAGs accumulates in one memory (personal data masked) — institutional know-how grows over time and survives staff turnover. You stay in full control: you decide, by permission, which teams and models can tap into it.
Permission matrix
Who writes, who reads — clearly
Access rules for each memory layer are defined up front; for corporate memory, privacy and permission control take center stage.
| Memory layer | Who writes? | Who reads? | Isolation & privacy |
|---|---|---|---|
| Model memory | Anyone using that model (if enabled on it) | The same model (optionally per user) | Isolated per model; can be split per user |
| Personal (seat) memory | Only you — across ALL models you use | Only you | Encrypted, isolated and privacy-compliant; control and deletion with the user |
| Corporate memory Highlighted | If on, ALL traffic: every model, chat and RAG (personal data masked) | Only opted-in seats OR models granted 'corporate access' (off by default) | Locked to the tenant; accessed by permission (Corporate Memory — View/Manage); PII masked before writing |
On SaaS, memory vector storage is billed from your wallet (GB-month); on-prem it stays within your organization. Deletion (right to be forgotten) is supported on every layer.
Free · no membership required
AI Akademi
127 modules, 1400+ units for businesses — from AI literacy to industry-specific expertise, step by step and hands-on. No sign-up needed.
One subscription
One account, endless possibilities
No need to sign up to every AI provider, register a separate card and track different invoices each month. In Operayn, one subscription unlocks all models.
- All models in one panel, one invoice
- Pay as you go or a fixed plan — choose what fits
- Real-time balance, quota and spend control
- Full e-invoice compliance
Monthly usageTransparent
gpt-4o%80claude%55gemini%35özel-model%20See every model's usage and cost from a single screen.
custom-legal-model
custom-ecommerce-assistant
custom-health-model
Models tailored to your industry and trained on your data — from the same API.
Your AI operator
Models we train for your industry — we sell solutions, not APIs
When general models aren't enough, WE train models on your business and data. We're not just an interface: we understand the problem, train the model, host it, scale it and deliver it working — you get the outcome.
- Accuracy tailored to your industry/data
- The same simple API as general models
- Hosting & scale on us — zero maintenance
Enterprise & On-Premise
Deploy on your own infrastructure — full control, full data sovereignty
Don't want your data to leave your organization? Deploy Operayn on your own servers, your own data center or your private cloud. The same powerful platform — gateway, RAG, AI Studio, chat — this time entirely under your control, dedicated to a single company and able to run fully isolated from the outside world.
On your own servers
The whole platform runs on your hardware; data, documents and API keys never leave your organization.
Single-tenant, dedicated
Not a multi-tenant cloud; a single, fully isolated installation dedicated only to your organization.
Air-gapped capable
Can run on networks isolated from the outside world. Build a fully offline AI stack with local models.
Enterprise SSO (OIDC)
Connect to your existing identity provider (Azure AD, Keycloak, Okta…) via OIDC; users sign in with their corporate accounts.
Easy Docker deployment
Containerized, multi-stage deployment. Comes up with a single command; scaling and updates with standard tooling.
Full KVKK / GDPR compliance
Data sovereignty is yours. Encrypted secrets, audit logs, role-based access and enforced 2FA — all within your boundaries.
Your providers, your models
Bring your own API keys and model endpoints; use local/enterprise models (vLLM, Ollama, private endpoints) from the same gateway.
No usage metering, no surprise bills
On-prem has no pay-as-you-go; a flat annual license. Let your team use it without limits while costs stay predictable.
Ideal for regulated industries (finance, healthcare, public sector, defense) and large enterprises. Talk to our team about deployment, licensing and integration.
Request an enterprise consultationHow it works
In three steps, within minutes
Step 1
Sign up free
Create your account. No card required; ready in minutes.
Step 2
Get your API key
Generate your opn- key from the panel and pick your plan.
Step 3
Connect your app
Point base_url to Operayn, set the model name and send your first request.
The difference
One Operayn instead of juggling many
| Separate providers | Operayn | |
|---|---|---|
| Access to all models | A separate account for each | All in one |
| Billing | Separate and messy | All in one |
| If a provider goes down | Your service stops | All in one |
| Batch processing | Usually missing / manual | All in one |
| Custom-trained model | Needs separate infrastructure | All in one |
| Usage/spend tracking | Scattered | All in one |
FAQ
Frequently asked questions
Yes. From a single account you call different models just by changing the model name; everything combines into one invoice.
No. Operayn is OpenAI-compatible; you only change base_url and api_key. Streaming and batch requests are supported too.
You pay as you go (token-based). You can pre-load a wallet or choose a subscription/quota plan. You see your spend on the dashboard for every request.
Automatic failover kicks in and your request is routed to a backup endpoint — your app keeps working.
You submit a request; a model is trained on your business/data, hosted, and served from the same API under a dedicated model name.
Secrets are stored encrypted, each customer is kept in an isolated database, and there's 2FA and audit logging. Your data is processed in line with data-protection law.
All AI models, waiting for you in one subscription
No setup, no card — create your account and send your first request in minutes.