Connect your workload to the trial.
Use an OpenAI-compatible client to test DeepSeek-V4.1-Flash, inspect usage and review the results before discussing your deployment.
On this page
Connect
Use the exact model ID below. OpenAI compatibility describes the request and response format, not identical model behavior or support for every optional parameter.
- Base URL
https://api.tiyuvta.ai/v1- Model ID
deepseek-ai/deepseek-v4.1-flash- Request
POST /v1/chat/completions- Authentication
Authorization: Bearer <your key>
An unknown model ID is rejected rather than silently substituted.
Streaming
The model can reason before answering. Reasoning is separate from answer content and is billed as output. Leave enough output budget for both.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.tiyuvta.ai/v1",
apiKey: process.env.TIYUVTA_KEY,
});
const stream = await client.chat.completions.create({
model: "deepseek-ai/deepseek-v4.1-flash",
messages: [{ role: "user", content: "Explain the trade-offs in this deployment plan." }],
stream: true,
stream_options: { include_usage: true },
});
for await (const chunk of stream) {
if (chunk.error) throw new Error(chunk.error.message);
const delta = chunk.choices[0]?.delta ?? {};
const reasoning = delta.reasoning_content ?? delta.reasoning;
if (reasoning) process.stderr.write(reasoning);
if (delta.content) process.stdout.write(delta.content);
if (chunk.usage) console.error(chunk.usage);
}Trial credit
Add prepaid trial credit in the console. Credit packs start at $2.00. Deployment engineering and support are agreed separately.
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.
Inspect usage in the console. Treat an absent or null cached-token detail as no reported cached tokens. Read the billing and refund terms before paying.
Errors
400- Invalid request, model or parameter
401- Authentication
403- Revoked or restricted access
402- Credit or spending limit
429- Admission or rate limit
5xx- Service failure
Correct the problem named by the error. Honor Retry-After when present and use bounded backoff for retryable failures. Do not retry indefinitely.
If the problem persists, contact our team with the request identifier and time, not your key.
Keys and data
Manage keys · Data handling · Privacy · DPA
Review the trial data conditions before sending sensitive information. Purchased and promotional credit have different conditions; any training opt-in is described in the privacy policy.