OpenAI-compatible API
Chat completions, embeddings and model listing follow the specification your team already knows. Migrate by changing one environment variable.
Every frontier model.
One gateway, one invoice.
MCLY Technology routes your prompts to OpenAI, Anthropic, Google, Meta, DeepSeek and more through a single OpenAI-compatible endpoint — and rents the GPU capacity to run your own weights when you need it.
No credit card on file required for a quote · Enterprise invoicing available
$ curl https://api.mclytechnology.com/v1/chat/completions \ -H "Authorization: Bearer $MCLY_API_KEY" \ -d '{"model":"gpt-4o","messages":[ {"role":"user","content":"Summarize this ticket"}]}' { "id": "chatcmpl-9xK2...", "model": "gpt-4o", "provider_routed": "openai/global", "usage": { "total_tokens": 412 }, "latency_ms": 384 } $
Providers
OpenAI
Anthropic
Google
Meta
DeepSeek
Mistral
Qwen
OpenAI
Anthropic
Google
Meta
DeepSeek
Mistral
Qwen
15+
Models on one endpoint
7
Upstream providers
99.9%
Monthly uptime target
<400ms
Median routing overhead
01 / UNIFIED ACCESS
Replace six SDKs with one
Point your existing client at our base URL and swap the key. Retries, fallbacks and provider failover are handled upstream, so a single provider incident never takes your product down.
Smart provider routing
Requests land on the healthiest upstream for that model. Automatic retry across regions keeps tail latency and error rates low without code changes.
Usage you can audit
Every request is metered per model with token-level detail. Track burn-down per key, per project, per team — and forecast next month's spend.
02 / ARCHITECTURE
How a request travels
YOUR APP
OpenAI SDK · any language
→
MCLY GATEWAY
auth · metering · failover
→
openai
anthropic
google · meta · deepseek
03 / PLANS
Compare all plans
Start with credits, scale to compute
Token packs spend across every model we route. Compute plans reserve dedicated GPU capacity by the month. Both are invoiced in USD.
Growth Pack
Most Popular
For production workloads that need premium models and priority routing.
$499.00
/ one-time
- $625 in API credits (25% bonus included)
- Standard + premium model pools
- Priority routing & automatic failover
- Credits valid for 24 months
Scale Pack
For high-volume platforms running millions of requests per month.
$1,999.00
/ one-time
- $2,700 in API credits (35% bonus included)
- All models including the flagship tier
- Dedicated capacity & lowest-latency routing
- 99.9% uptime SLA
H100 Node
Most Popular
Flagship Hopper performance for frontier training and low-latency serving.
$1,499.00
/ per month
- Dedicated NVIDIA H100 80 GB (SXM)
- 48 vCPU · 512 GB RAM · 4 TB NVMe
- NVLink · InfiniBand ready
- 99.9% uptime SLA
Running your own weights? Dedicated H100, A100 and RTX 4090 instances are available by the
month with root access, persistent volumes and hourly burst options —
see compute pricing.
04 / MODEL CATALOG
Browse all models
A catalogue that keeps up
Flagship and cost-efficient models side by side — switch with the model parameter, no new contract.
GPT-4o
128K
OpenAI
OpenAI's flagship multimodal model for text, vision and audio.
o1
200K
OpenAI
Reasoning model that thinks step by step on hard problems.
Claude Opus 4
200K
Anthropic
Anthropic's most capable model for complex agentic work.
Claude Sonnet 4
200K
Anthropic
Balanced frontier intelligence and speed for production.
Gemini 2.5 Pro
1M
Google's flagship model with a 1M-token context window.
Llama 4 Maverick
1M
Meta
Meta's open-weight MoE model with 1M-token context.
DeepSeek R1
64K
DeepSeek
Open reasoning model competitive with frontier labs.
05 / ENTERPRISE
Invoicing your finance team will accept
Larger customers are billed by bank transfer or corporate card. Email us your order number and our team returns bank details or a secure payment link the same business day, together with a proper invoice referencing MCLY TECHNOLOGY INC.
Request a quote
06 / SUPPORT
Engineers on the other end
Integration questions, quota increases, private deployments and custom model hosting — one inbox, one business day, no ticket roulette. Availability and response targets are published in our SLA.
[email protected]
07 / FAQ
Before you ask
Do I need to change my code to switch models?
No. Every model is exposed through the same OpenAI-compatible schema.
Change the model field and the gateway routes to that provider.
How are token credits consumed?
Credits are drawn down per request based on the upstream provider's token
accounting, at the rate card attached to your plan. Unused credits stay on the account until
the validity window in your plan ends.
Can I pay by bank transfer?
Yes — larger orders are settled by bank transfer or corporate card.
Place the order, then email [email protected]
with your order number and we will send the details.
Which GPU instances are available?
RTX 4090, A100 and H100 instances, from single cards to eight-GPU clusters,
reserved monthly with root access. See the compute section on the pricing page.