OpenAI · Conversation

GPT-5.4 API

GPT-5.4 is a text model by OpenAI. On Evrona you call it with the same API key and balance as every other model in the catalog. Its context window takes up to 1,050,000 input tokens. Requests go to the OpenAI-compatible /v1/chat/completions or the Anthropic-compatible /v1/messages; streaming is supported, and function calling works where the model supports it.

Model ID: openai/gpt-5.4
Developer: OpenAI
Type: Conversation
Context: 1,050,000 tokens

GPT-5.4 pricing

Billed for	Price
Input tokens, per 1M	$0.28
Output tokens, per 1M	$1.97
Typical request: 2,000 input and 500 output tokens	$0.0015

Accounting inside the service is in USD; EUR prices are recalculated at the current rate and may change.

GPT-5.4 request example

Use https://evrona.ru/v1 as the base URL and pass openai/gpt-5.4 in the model field. You create the key in your account after signing up.

curl request

curl

curl https://evrona.ru/v1/chat/completions \
  -H "Authorization: Bearer ev-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.4",
    "messages": [
      {
        "role": "user",
        "content": "Hi! Tell me about yourself in two sentences."
      }
    ]
  }'

Python with the official OpenAI SDK

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://evrona.ru/v1",
    api_key="ev-YOUR_KEY",
)

completion = client.chat.completions.create(
    model="openai/gpt-5.4",
    messages=[
        {
            "role": "user",
            "content": "Hi! Tell me about yourself in two sentences."
        }
    ],
)
print(completion.choices[0].message.content)

Claude Code and Anthropic-protocol clients

Clients that speak the Anthropic protocol take the address without /v1: ANTHROPIC_BASE_URL=https://evrona.ru. The model ID stays the same.

macOS / Linux

export ANTHROPIC_BASE_URL=https://evrona.ru
export ANTHROPIC_AUTH_TOKEN=ev-YOUR_KEY
export ANTHROPIC_MODEL=openai/gpt-5.4
claude

Where to use GPT-5.4

Step-by-step guides for tools where you can pick this model: