Hosted open models

Open-model inference, measured on the live endpoint.

The figure beside this line was measured on the endpoint you call, and the command under it is that measurement. tiyuvta runs its own engine and serving stack: nothing here is resold capacity. Every model serves its full native context at one price per token, prepaid, so a loop cannot produce an invoice.

Get an API key Read the quickstart

See every measurement and its conditions

Production volume.

Last measured

2026-08-25 UTC

Time to the first token you can read: 230 ms

Reproduce it

curl -N https://api.tiyuvta.ai/v1/chat/completions \
  -H "Authorization: Bearer $KEY" \
  -d '{"stream":true,
       "model":"ornith-ai/ornith-1.5-35b-a3b",
       "messages":[{"role":"user",
                    "content":"hi"}]}' \
  -o /dev/null \
  -w 'ttft %{time_starttransfer}s\n'

How each figure was measured, and where it does not hold.

5 models · one endpoint
open engine memra, Rust + CUDA, ours
public status status.tiyuvta.ai, per model
reproducible every figure carries its curl

Models available now

Measurements and model details

Whose number Benchmark scores under each model name are the maker’s own published figures. They are not ours and are not comparable between makers. Output speed is measured by us on the live endpoint. Those are observations, not service guarantees. Every figure, with its conditions.

Billing: No subscription, no minimum. Credit does not expire.
Switching: Standard APIs. No product-specific SDK to remove.

Make the first request

This request uses OpenAI Responses. Create a key, copy it, and replace the prompt. Tiyuvta does not require a product-specific SDK.

Agent client configurations · Integration directory

curl

curl https://api.tiyuvta.ai/v1/responses \
  -H "Authorization: Bearer $TIYUVTA_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"ornith-ai/ornith-1.5-35b-a3b","input":"Hello."}'

Your prompts stay yours

Prompts are processed in memory to answer the request and are never used to train, fine-tune, distil or evaluate a model. There is no sampling queue and no human review.

If you want to, you can turn training on for your own traffic and take 5% of your metered spend back as credit. It is one switch in the console, off unless you set it, and turning it off stops it for everything sent afterwards. The full data policy.

Operator and production traffic

Avi Fenesh builds memra and operates this service. For production traffic, email support@tiyuvta.ai with the model, expected concurrency, and token volume.