Turkey's Gateway to AI
Turkey's most comprehensive AI API platform. Access 690+ AI models with a single API and transparent Turkish Lira pricing.
690+
AI Models
82+
AI Providers
%99.9
Uptime SLA
₺ TL
Local Pricing
4
Modalities (Text/Image/Video/Music)
7/24
Turkish Support
Platform Features
The technical capabilities that set Onysoft AI Gateway apart
Multi-Modality Support
Text (chat), image generation (text-to-image, image-to-image), video (text-to-video, image-to-video, video-to-video), music and sound effect generation — all from a single platform. No separate integration needed to switch modalities: the same API key, the same endpoint logic.
Streaming & Async Task Management
Real-time streaming for chat responses via Server-Sent Events (SSE). A task_id-based async flow for video, image and music generation — ideal for long-running jobs with status polling and webhook support.
Partner (Reseller) Program
A dedicated partner panel for digital agencies and solution partners. Manage your own projects and customers, set a custom markup (profit margin), per-model token limits and separate API keys for each. Detailed earnings reports and profit analysis included.
Per-Model Token Limits
Define daily/monthly token and request limits per model for each API key. Provider-level limiting with wildcard pattern support (e.g. google/*). Option to block specific models entirely.
Smart Optimization Recommendations
Model-family-based optimization suggestions driven by usage analysis. "Cheaper equivalent model" recommendations — e.g. suggesting GPT-4o-mini instead of GPT-4o for savings of up to 96%. Estimated monthly savings calculation.
Playground (Comparison)
Compare multiple models side by side with the same prompt. Chat, comparison and generation modes. Quick testing, cost comparison and history viewing. Try models without writing any code.
Bring Your Own API Key (BYOK)
Manage your own OpenAI, Anthropic, Google and other API keys through Onysoft. Assign different provider keys per project. Onysoft acts as both a payment bridge and a multi-tenant management layer.
Detailed Request Logs
Request and response payloads (JSON) are stored for every API call. A full audit trail with token, cost, latency, status and IP address details. View and search from the admin panel.
Our Story
Developers and businesses in Turkey face serious barriers when accessing AI API services: foreign credit card requirements, high exchange rates, and invoicing and accounting headaches. Onysoft AI Gateway was founded to eliminate these problems.
Developed by Izmir-based Onysoft Veri Merkezi A.Ş., our platform unifies all the models of the world's leading AI providers — OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek — under a single API.
Our goal is simple: to give Turkish entrepreneurs, developers and organizations the easiest, most affordable and most secure access to AI technologies.
What We Offer
Access 690+ AI models through a single OpenAI-compatible API — text, video, image and music generation
Unified AI API
Reach GPT-5.2, Claude 4.6 Opus/Sonnet, Gemini 3.1 Pro, Llama 4, Mistral, DeepSeek V3.2 and 690+ models from a single integration point. Includes video (Veo 3.1, Sora 2, Kling, Runway, Wan), image (Flux 2, Imagen 4, Seedream, Ideogram V3, GPT Image 1.5, Grok Imagine) and music (Suno V4, ElevenLabs) generation models. Use your existing OpenAI SDK code unchanged — just swap the base_url parameter.
Pay in Turkish Lira
Transparent pricing in TRY based on the current TCMB exchange rate. Top up your balance easily by credit card or bank transfer. End accounting headaches with corporate e-invoice support. Fixed TRY prices, unaffected by exchange-rate swings.
High Performance & Low Latency
Minimal latency thanks to our optimized infrastructure. Real-time responses for chat models with SSE streaming support. A task_id-based polling mechanism for async video/image generation. 99.9% uptime guarantee.
KVKK-Compliant Security
Data processing policies aligned with Turkish regulations. All API traffic is transmitted over TLS 1.3. API keys are stored hashed, with brute-force-protected authentication. Enterprise-grade data security and 2FA support.
Detailed Usage Analytics
Track the input/output token usage, provider cost, customer cost, profit and performance of every API call in real time. Per-model, daily and monthly reports. A full audit trail with an archive of request/response payloads (JSON).
Get Started in 5 Minutes
Create a free account, get your API key and start using it right away. Comprehensive documentation plus ready-made Python, JavaScript/Node.js, PHP and cURL code examples. Full OpenAI SDK compatibility.
Advanced Control Mechanisms
Define markup rates per project/customer/key, per-model daily and monthly token limits, rate limiting (per minute/day), blocking of specific models, and an allowed_models whitelist. Give your own customers granular control through the partner panel.
Smart Optimization
Your usage data is analyzed to suggest cheaper alternatives within the same model family. Automatic recommendations like GPT-4o → GPT-4o-mini unlock monthly savings of up to 96%. Profit margin, provider cost breakdown and trend charts in your dashboard.
Playground — Try Without Writing Code
Try it instantly with the free, no-signup demo widget. Once signed in, access full Playground chat, comparison and generation modes. Send the same prompt to multiple models and compare the responses side by side.
BYOK — Bring Your Own Key
Manage your own OpenAI, Anthropic, Google AI Studio and other provider keys through Onysoft. Assign different provider keys per project. Onysoft provides both multi-tenant management and a Turkish Lira payment bridge.
Partner (Reseller) System
A white-label partner panel for digital agencies, software houses and solution partners. Open sub-accounts for your own customers, earn through markup, and get detailed earnings reports. Dedicated API key, balance and per-model limits for each customer.
24/7 Turkish Technical Support
Instant help in Turkish via email, phone and live support channels. Integration support, model selection guidance and price optimization advice. Written follow-up through support tickets.
Supported AI Providers
Access all the models of the world's leading AI companies from a single platform
OpenAI
GPT-5.2, GPT-5.1, GPT-4.1, GPT-4.1 Mini/Nano, o3, o4-mini, Sora 2 (video), Sora 2 Pro, GPT Image 1.5 (image)
Anthropic
Claude Opus 4.6, Claude Sonnet 4.6, Claude Haiku 3.5 — 200K context, tool use, vision
Gemini 3.1 Pro (1M context), Gemini 2.5 Flash, Veo 3.1 (video), Imagen 4 Ultra/Fast (image), Nano Banana
Meta
Llama 3.3 70B, Llama 4 Maverick, Llama 4 Scout — open weights, scales cheaply
Mistral
Mistral Large, Medium, Small, Nemo, Devstral, Pixtral (vision), Codestral
DeepSeek
DeepSeek V3.2, DeepSeek R1 (reasoning), DeepSeek Chat — cheap and powerful
xAI (Grok)
Grok 4, Grok 3, Grok Imagine (video/image)
Alibaba Qwen
Qwen 3 (14B-235B), Qwen Max, Qwen Plus, Qwen VL (vision)
Kling AI
Kling 2.5 Turbo, Kling 2.6, Kling 3.0 — high-quality t2v/i2v
Runway
Runway Gen-4 Turbo, Runway Aleph (v2v) — cinematic video generation
ByteDance
Seedream 4.5, Seedream 5.0 Lite — image generation and editing
Black Forest Labs
Flux 2 Pro, Flux 2 Flex — high-resolution image generation
Ideogram
Ideogram V3, Character, V3 Edit, V3 Remix — text-render-focused image generation
Alibaba Wan
Wan 2.5, Wan 2.6 (t2v/i2v/v2v), Wan 2.7 Image Pro
Hailuo
Hailuo 02, Hailuo 2.3 Pro/Standard (i2v)
Suno
Suno V4 — song generation (instrumental/vocal)
ElevenLabs
Text-to-Speech (Multilingual v2, Turbo 2.5), Sound Effects v2, Audio Isolation
Others
AI21 (Jamba), Cohere, Perplexity Sonar, Recraft, Topaz, NVIDIA, Together AI, Fireworks, Groq, Aion Labs and many more
How It Works
Start accessing the AI API in 3 steps
Create a Free Account
Sign up in seconds with your email or Google / GitHub account. No credit card required.
Top Up Balance & Get an API Key
Top up your balance in TRY by credit card or bank transfer. Create your API key from the dashboard.
Make Your API Calls
Use our OpenAI SDK-compatible API to instantly access GPT-5, Claude 4.6, Gemini 3.1 and 690+ models. Video, image and music generation included. Just change the base URL.
Who Is It For?
For everyone who needs an AI API
Software Developers
Individual and freelance developers building AI applications like chatbots, AI assistants, content generators and coding helpers.
Startups & Tech Companies
SaaS companies, fintech startups and e-commerce platforms integrating AI into their products. Corporate invoicing and bulk usage advantages.
Researchers & Academics
Universities and research institutions working on NLP, natural language processing, text analysis and AI research.
Agencies & Solution Partners
Digital agencies and software houses delivering AI solutions to their clients. Open and manage sub-accounts for your own customers through our partner program.
Technical Architecture
The technologies and flow logic underpinning the platform
api API Endpoint Structure
Manage all modalities through two main endpoints:
POST /v1/chat/completions— Text/chat/multimodal models (synchronous, streaming supported)POST /v1/video/generate— Video, image and music generation (async, task-based)GET /v1/video/status/{task_id}— Async task status queryGET /v1/models— Available models and markup-applied pricesGET /v1/balance— Real-time balance and summaryGET /v1/usage— Detailed usage logs (filtering, pagination)
vpn_key Authentication Layers
The system supports 4 different API key types, all in the same Bearer token format (sk-ony-):
- User Key — Individual end user created from the dashboard (TRY balance)
- Project Key — Organization projects (TRY balance, custom markup)
- Partner Project Key — Partner projects (USD balance, partner markup)
- Partner Customer Key — Partner customers (TRY balance, customer markup)
calculate Pricing and Markup Flow
The system uses 3-tier pricing: provider_cost (what we pay the provider, in USD) → customer_cost (markup-applied customer price, in USD) → cost_tl (in TRY at the current TCMB rate). The markup rate can be customized at the project/customer/key level. Token-based models are charged separately for input/output, while generation-based models (video/image/music) are charged per generation.
shield_lock Security Layers
- API keys are stored hashed with SHA-256
- Brute-force protection: IP-based failed-attempt counter, 15-minute block
- Rate limiting: 60 requests per minute, 10,000 requests per day (default, increases by plan)
- 5 requests per 24 hours on free (:free) models
- Per-model daily/monthly token limits (partner feature)
- 2FA support (TOTP + WebAuthn) — for admin and partner accounts
- Request/response content policy filtering
- End-to-end encryption with TLS 1.3
- KVKK- and GDPR-compliant data processing
dns Infrastructure
High-performance backend built on PHP 8.2+. MySQL 8 primary database plus file-based model cache (5-minute TTL). 400+ text/chat models via OpenRouter, 200+ video/image/music models via KieAI. Real-time USD/TRY rate through TCMB API integration. Windows Server infrastructure running on Microsoft IIS, with automatic backup and failover support.
Use Cases
What can you build with Onysoft AI Gateway?
Customer Service Chatbot
Build smart assistants powered by GPT-5 or Claude 4.6 that run 24/7, understand Turkish and provide product information. Integrate with your own knowledge base using RAG (Retrieval Augmented Generation).
Content Generation and Copywriting
Blog posts, product descriptions, email campaigns, social media posts. SEO optimization, headline variations, meta descriptions. Batch API support for bulk content generation.
Coding Assistant and Developer Tools
Code generation, debugging, code review and documentation with Claude Sonnet 4.6, GPT-5.2 and DeepSeek V3.2. Code-specialist models like Devstral and Codestral. Response times suitable for IDE extensions.
Image Generation and Editing
Marketing visuals, product photo mockups, social media content, illustrations. High-quality images with Flux 2 Pro, Imagen 4 and GPT Image 1.5. Edit existing images with image-to-image.
Video Ads and Social Media
Short video ads, product showcases and TikTok/Reels content with Veo 3.1, Sora 2 and Kling 3.0. Text-to-video, image-to-video and video-to-video conversion. 5-10 second 1080p/4K video output.
Music and Sound Effect Generation
Song generation (instrumental or with vocals) with Suno V4. Text-to-speech with ElevenLabs (29+ languages including Turkish). Podcast intros, sound effects and game music generation.
Translation and Localization
Book, document and software localization with large context-window models (Gemini 3.1 Pro 1M tokens, Claude 200K). Turkish-English and 100+ language pairs. Consistency-preserving translation.
Data Analysis and Reporting
AI agents that analyze JSON/CSV data, produce summaries and generate insights. Function calling and structured output support. Real-time data querying with Claude and GPT tool use.
Education and Research
Academic research, literature review and summarization. An affordable test environment for NLP projects. University and student discounts available.
Frequently Asked Questions
What is Onysoft AI Gateway?
Which AI models do you support?
Can I use my existing OpenAI integration?
How does pricing work?
What is the difference between the Chat API and the Video/Image API?
How do image-to-image models work?
Can I define per-model token limits?
How do smart optimization recommendations work?
What is the Partner (Reseller) program?
Can I use my own API key (OpenAI, Claude, etc.)? (BYOK)
What is Playground?
What is the rate limit? How much?
Are you KVKK compliant?
Do you provide technical support?
Can I get an invoice?
How often are new models added?
Corporate Information
Trade Name
ONYSOFT VERİ MERKEZİ SİSTEMLERİ YAZILIM SANAYİ VE TİCARET ANONİM ŞİRKETİ
İTOB OSB MAH. 10032 SK. NO: 2 İÇ KAPI NO: Z13
MENDERES / İZMİR
info@onysoft.com
0850 302 3021
api.onysoft.com
Explore AI in Turkish Lira
Access 690+ AI models right away. Create a free account, top up in TRY, and integrate within minutes.