Infersia
Sign inGet an API key
ModelsPricingDocsStatusTransparency

Quickstart

Two changes to existing OpenAI code, and you’re running.

1. Get a key

  1. Create an account — you get $1 of credit, no card required.
  2. Go to Dashboard → API keys and create one. The full key is shown once; we store only a hash of it and genuinely cannot recover it later.
  3. Export it: export INFERSIA_API_KEY=isk-v1-…

2. Point your client at us

Python

chat.pypython
from openai import OpenAI
import os

client = OpenAI(
    base_url="https://api.infersia.com/v1",
    api_key=os.environ["INFERSIA_API_KEY"],
)

response = client.chat.completions.create(
    model="qwen/qwen3-8b",
    messages=[{"role": "user", "content": "Explain KV caching in two sentences."}],
)

print(response.choices[0].message.content)
print(f"Cost: ${response.usage.cost:.8f}")

TypeScript

chat.tstypescript
import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://api.infersia.com/v1',
  apiKey: process.env.INFERSIA_API_KEY,
});

const stream = await client.chat.completions.create({
  model: 'qwen/qwen3-8b',
  messages: [{ role: 'user', content: 'Explain KV caching.' }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? '');
}

curl

curlbash
curl https://api.infersia.com/v1/chat/completions \
  -H "Authorization: Bearer $INFERSIA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen/qwen3-8b",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": true
  }'

3. Check what you got

Every response carries headers describing exactly what served it. This is the part that’s hard to get from most providers:

response headershttp
x-request-id:              isr_01KYGHW7S1GTR2AVE40C
x-infersia-provider:      infersia
x-infersia-quantization:  awq-int4
x-infersia-cost-usd:      0.000001950
x-ratelimit-remaining-requests: 199
x-ratelimit-remaining-tokens:   399987

You can assert on x-infersia-quantization in your own test suite. If we ever change the precision behind a model you depend on, your tests catch it — you don’t have to take a page on our website at its word.

Requests are billed from the inference engine’s own token counts, not from a tokeniser we run separately. Your invoice and the usage object in your response are the same number.

Next

Read about model routing to pin a provider or use the free tier, errors and limits to handle 429s and 402s correctly, or the integration guides if you’re wiring up a coding tool rather than writing code directly.

Quickstart · Infersia