Skip to content
PrByCode
Sign inGet API Key
Get API Key
OpenAI-compatible API

One API for every model.

Route requests across leading AI models through one compatible endpoint. Keep your client, change the base URL, and manage keys, usage and fallbacks in a single console.

Get API KeyView Models
https://api.prbycode.com/v1
No client rewritePer-request metersFallback routing
POST /v1/chat/completions
POSThttps://api.prbycode.com/v1/chat/completions
{
  "model": "prbycode/model-name",
  "messages": [
    { "role": "user", "content": "Hello" }
  ],
  "stream": true
}
Status200
Latency842 ms
Tokens1,284
Provideranthropic
OpenAI compatibleDrop-in base URL
Anthropic compatible/v1/messages support
StreamingServer-sent events
Large contextUp to 1M tokens
Unified endpointOne key, one URL

Popular models

Frontier and fast models behind one key.

Every model speaks the same request shape. Switch by changing a single field.

Full catalog
ProviderModelContextInput / 1MOutput / 1M
AnthropicClaude Sonnet 5claude-sonnet-51M$3.00$15.00Details
Zhipu AIGLM 5.3 Flashglm-5.3-flash256K$0.25$1.10Details
Zhipu AIGLM 5.2glm-5.2256K$0.60$2.20Details
DeepSeekDeepSeek V4 Prodeepseek-v4-pro256K$0.90$3.40Details

Why PrByCode

Built like infrastructure, not a marketing page.

One key, every model

Issue scoped keys from the console and point any client at one base URL.

Usage you can audit

Requests, tokens and spend per model and per key, without guessing.

Failures that recover

Retry and fallback chains move a request to a healthy route automatically.

EndpointsOpenAI chat completions and Anthropic messages
StreamingServer-sent events on every routable model
ObservabilityRequest-level logs with full payloads

Routing and reliability

A single path, with a fallback behind it.

Configure a default model and an ordered fallback chain. Failed calls retry, unavailable routes are skipped, and every hop is recorded in your logs.

Your applicationAny OpenAI-compatible client
https://api.prbycode.com/v1Single authenticated entry point
PrByCode gatewayRequest validation and metering
Model routingDefault model and fallback chain
Provider inferenceUpstream model execution

API example

Two lines to change.

Keep the SDK you already depend on.

curl https://api.prbycode.com/v1/chat/completions \
  -H "Authorization: Bearer $PRBYCODE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3-flash",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'
API keyPRBYCODE_API_KEY
Base URLhttps://api.prbycode.com/v1

Console

Is the API working, and what is it costing?

Requests, tokens, spend and success rate per period, with the failing requests one click away.

Open the console

Pricing

Pay for what you use.

Model usage is metered per token. Access periods below cover the console and the rate limits on your keys.

Full pricing
Access periodPriceCoversBest for
Daily$3 / 24 hoursConsole, keys, logs and routingFocused access for short sessions and evaluation.Details
WeeklyMost common$8 / 7 daysConsole, keys, logs and routingFlexible access for active builds and sprints.Details
Monthly$22 / 30 daysConsole, keys, logs and routingConsistent access for regular development.Details
Yearly$199 / 365 daysConsole, keys, logs and routingLong-term access at the lowest daily cost.Details

Quickstart

Your first routed request in three steps.

No SDK migration and no new request shape.

Get API KeyRead the docs
  1. 1
    Create an API key

    Generate a key in the console and store it as an environment variable.

  2. 2
    Copy the endpoint

    Point your client at https://api.prbycode.com/v1.

  3. 3
    Send a request

    Keep the request shape you already use and pick any model ID from the catalog.