Test our deployment with your own work.
Add prepaid trial credit and test DeepSeek-V4.1-Flash on our reference deployment. Then talk to us about your own deployment.
Trial credit starts at $2.00. It pays for endpoint usage, not deployment engineering or support.
Review the trial terms, data handling and refund policy before paying.
Connect an OpenAI-compatible client
- Base URL
https://api.tiyuvta.ai/v1- Model ID
deepseek-ai/deepseek-v4.1-flash- Request
POST /v1/chat/completions
How it runs
Sign in
Open the trial console.
Add credit and create a key
Choose a credit pack and store your key securely.
Run your evaluation
Send representative requests and inspect results and usage.
What the trial tells you
This tests our reference deployment, not the performance of different hardware. Your deployment is measured separately on your model, hardware and workload.
Check trial statusTrial metering
Per 1,000,000 tokens: $0.30 uncached input; $0.006 reported cached input; $1.20 output, including reasoning. The cached rate applies only to tokens reported as cached. Repeating a prefix does not guarantee a cache hit.
Read metering detailsRequest a different model test
Running another model or a different hardware class? Tell us what you run and request a one-hour test window. We confirm the model, hardware and timing before deploying it for the session.
For a confirmed test, we agree the measurements and review the results with you. The receipt names the model, hardware, settings and results.
Policies
Prefer to talk first? Book a conversation
The calendar books a planning conversation, not a confirmed model test. You can use either route.