Migrate from DeepInfra to Railwail

TL;DR — Switch in Under 5 Minutes

• DeepInfra is already OpenAI-compatible — change baseURL only
• All major open models mirrored: Llama 3.3, Mixtral, Qwen 2.5, DeepSeek V3 / R1
• Image, audio, embeddings — all supported
• EU-hosted endpoint, EUR billing
• Plus closed-source Claude, GPT-4o, Gemini behind the same key

Why Move Off DeepInfra?

DeepInfra is competitive on price for hosted open-source models. The trade-offs: US-default routing, USD-only billing, no closed-source frontier models, and a smaller image / audio catalog than Railwail. Migrating is essentially a baseURL change.

Step 1 — Get a Railwail API Key

Sign up at railwail.com and generate a key.

Step 2 — Change Base URL

TypeScript / JavaScript

Before (DeepInfra):

import OpenAI from "openai";

const di = new OpenAI({
  apiKey: process.env.DEEPINFRA_API_KEY,
  baseURL: "https://api.deepinfra.com/v1/openai",
});

const res = await di.chat.completions.create({
  model: "meta-llama/Meta-Llama-3.3-70B-Instruct",
  messages: [{ role: "user", content: "Hello" }],
});

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.RAILWAIL_API_KEY,
  baseURL: "https://api.railwail.com/v1",
});

const res = await client.chat.completions.create({
  model: "llama-3.3-70b-instruct",
  messages: [{ role: "user", content: "Hello" }],
});

Python

from openai import OpenAI

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

resp = client.chat.completions.create(
    model="llama-3.3-70b-instruct",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

cURL

curl https://api.railwail.com/v1/chat/completions \
  -H "Authorization: Bearer $RAILWAIL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama-3.3-70b-instruct",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

API Endpoint Mapping

Model Mapping

Pricing Comparison (per 1M tokens, May 2026)

Why Railwail Over DeepInfra

• EU billing in EUR with VAT receipts
• Frankfurt-region hosting for low EU latency
• Same OpenAI-compatible API
• Adds closed-source models (Claude, GPT-4o, Gemini)
• Built-in playground at railwail.com/models
• Per-key rate limits and spend caps

FAQ

What about DeepInfra's deploy-your-own-model feature?

Custom model deployments are not yet supported on Railwail. Keep them on DeepInfra and use Railwail for the standard catalog.

Does the embedding API work the same?

Yes. POST /v1/embeddings accepts identical request shapes for BAAI/bge, Sentence Transformers and Cohere Embed models.

What is the throughput per request?

Comparable to DeepInfra for the same model. Railwail uses vLLM with continuous batching on EU hardware.

Are streaming responses supported?

Yes. Standard OpenAI SSE stream chunks.

What about quantised model variants?

FP8 / int4 variants are exposed with -fp8 or -int4 suffix in the model slug when available.

Next Steps

• Sign up at railwail.com
• Generate an API key
• Update baseURL to https://api.railwail.com/v1
• Switch model IDs to Railwail short slugs
• Read the reference at railwail.com/docs
• Compare pricing at railwail.com/pricing