API for developers
Access leading
AI models.
Connect models to your applications with clear pricing and OpenAI and Anthropic compatible APIs.
- Billing
- usage based
- Formats
- OpenAI · Anthropic
- Capabilities
- Text, reasoning, tools
from openai import OpenAI
client = OpenAI(
max_retries=0,
api_key="YOUR_ASTRA_API_KEY",
base_url="https://platform.astragpt.net/api/v1",
)
response = client.chat.completions.create(
model="openai/gpt-6-luna",
messages=[{"role": "user", "content": "Hello!"}],
max_tokens=512,
)
print(response.choices[0].message.content)curl https://platform.astragpt.net/api/v1/chat/completions \
-H "Authorization: Bearer $ASTRA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-6-luna",
"messages": [{"role":"user","content":"Hello!"}],
"max_tokens": 512
}'platform.astragpt.net/api/v1Catalog
Models and pricing
Input and output token prices per 1 million. See the catalog for full conditions.
| Model | Input | Output | Context |
|---|---|---|---|
| GPT-6 Astra Proopenai/gpt-6-astra-pro | 1 400 ₽ | 7 000 ₽ | 922 000 |
| Claude Opus 5.5anthropic/claude-opus-5.5 | 560 ₽ | 2 800 ₽ | 1 000 000 |
| Gemini 4 Argongoogle/gemini-4-argon | 280 ₽ | 1 400 ₽ | 1 048 576 |
| Kimi K3.1 Promoonshotai/kimi-k3.1-pro | 280 ₽ | 1 400 ₽ | 1 048 576 |
Getting started
From API key to first response
The console tracks spending and key limits. A key secret is shown once, so save it immediately.
- 01
Log in with AstraGPT
Use the same account with a separate API wallet.
- 02
Create an API key
Optionally set an expiry date and allowed models.
- 03
Set the base URL
Use Chat Completions, stateless Responses, or Anthropic Messages.
Transparent pricing
Pay for what you use
Connect models to your applications and pay for actual token usage. Each model’s rates are listed in the catalog.
Verified saving: On 24 October 2026, AstraAPI’s base input and output prices for GPT‑6.1 Sol were 30% lower than GPTunneL and AI Tunnel: RUB 280 / 1,400 versus RUB 400 / 2,000 per 1 million tokens. The comparison excludes cache, long-context tiers, and personal discounts.