AI Gateway — One endpoint, every LLM

One endpoint.Every model. Yours.

BYOK across every major LLM provider. Auto-fallback. Orchestrator routing. Prompt-injection guard. PII redaction. Drop-in compatible with any OpenAI SDK.

First month free on paid plans · 0% provider markup · Forever-free tier · BYOK

Drop-in. Not rewrite.

Change one line. Keep the rest.

Every OpenAI SDK already speaks the gateway. Streaming, tool-calling, JSON mode, audio — works unchanged. Switch model to any provider on your account.

  • Compatible with openai-node, openai-python, ai-sdk, LangChain, LlamaIndex
  • SSE streaming, tools, function calling, JSON mode
  • Per-request override of provider, key and timeout
drop-in.ts
import OpenAI from "openai";

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

const res = await client.chat.completions.create({
  model: "dymo/auto",          // orchestrator picks the best provider
  messages: [{ role: "user", content: "Plan a launch in 5 bullets." }]
});

15 providers. One key. Bring your own.

OpenAIStandardAnthropicStandardGoogle AI StudioStandardGroqStandardMistralStandardxAI (Grok)StandardOpenRouterStandardCerebrasStandardAlibaba Cloud (Qwen)StandardMiniMaxStandardOpenCodeCommunityCAMB.AICommunityElevenLabsCommunityNovita AICommunityNVIDIACommunity
OpenAIStandardAnthropicStandardGoogle AI StudioStandardGroqStandardMistralStandardxAI (Grok)StandardOpenRouterStandardCerebrasStandardAlibaba Cloud (Qwen)StandardMiniMaxStandardOpenCodeCommunityCAMB.AICommunityElevenLabsCommunityNovita AICommunityNVIDIACommunity
OpenAIStandardAnthropicStandardGoogle AI StudioStandardGroqStandardMistralStandardxAI (Grok)StandardOpenRouterStandardCerebrasStandardAlibaba Cloud (Qwen)StandardMiniMaxStandardOpenCodeCommunityCAMB.AICommunityElevenLabsCommunityNovita AICommunityNVIDIACommunity
OpenAIStandardAnthropicStandardGoogle AI StudioStandardGroqStandardMistralStandardxAI (Grok)StandardOpenRouterStandardCerebrasStandardAlibaba Cloud (Qwen)StandardMiniMaxStandardOpenCodeCommunityCAMB.AICommunityElevenLabsCommunityNovita AICommunityNVIDIACommunity

Everything you would have built. Skipped.

Routing, fallback, observability, safety. The boring infra you don't want to own — and the safety surface you absolutely should.

OpenAI-compatible. Drop-in.

Keep every line of your OpenAI SDK code. Change the baseURL — that's it. Streaming, tools, function calling, JSON mode all work.

BYOK across 14 providers

Bring your own keys. You pay providers directly — Dymo charges only the subscription. No hidden token markup.

Automatic fallback

Provider down? Key rate-limited? Model deprecated? Traffic routes around the failure. Your users never see the 5xx.

Dymo Orchestrator

Map a task profile to the cheapest model that meets the bar. Cheap for cheap, smart for smart, automatically.

Prompt-injection guard

Heuristic and strict modes. Block jailbreaks, indirect injection and tool-call hijacking before the model sees them.

PII / data-leak guard

Redact emails, IDs, IBANs, custom regex — both ways. Your prompts and completions stay clean.

Audio: TTS + STT

Text-to-speech and speech-to-text across every supported audio provider. Same SDK. Same key.

Custom model aliases

Map dymo/fast or my-team/cheap to any upstream model. Swap the underlying model without redeploying clients.

Cost & usage analytics

Estimated cost per model, daily and weekly charts, breakdowns by alias. Catch a runaway prompt before billing does.

Built for production

No lock-in. No markup. No 5xx.

15
Providers supported
0%
Token markup
Auto
Fallback retries

Why teams switch to Dymo Gateway.

Avoid lock-in

Switch providers in a config flag. Negotiate against incumbents from a position of strength.

Survive provider outages

Auto-fallback turns a competitor's bad day into a non-event for your users.

Cut LLM spend

Route trivial calls to cheap models. Reserve frontier models for the queries that need them.

Ship safely

Injection guard + PII redaction before your prompts ever leave your perimeter. SOC-friendly logs included.

Build vs buy vs Dymo.

Honest table. The math usually picks itself.

CapabilityDymo GatewayBuild it yourselfOther gateways
Setup timeChange baseURLWeeksDays
BYOKSometimes
Provider markup0%0%5–20%
Auto-fallbackBuild it
Orchestrator routingLimited
Prompt-injection guardHeuristic + StrictAdd-on
PII redactionAdd-on
Audio (TTS / STT)Rare
OpenAI-compatiblePer-provider

Free to start. Priced for production.

Free tier forever. First month free on Hobby and Pro — no commitment. BYOK, no token markup. Pay providers directly; Dymo charges only the subscription.

Free

$0

500 req/mo · 50 RPM · 5 providers · Multi-provider fallback

Hobby

Freefirst month

then $20/mo

25k req/mo · 500 RPM · Orchestrator · Injection guard (heuristic)

Pro · Recommended

Freefirst month

then $79/mo

1M req/mo · 1.5k RPM · Audio · PII guard · Full safety stack

Enterprise

Custom

SSO · KMS · 99.99% SLA · Custom DLP · Dedicated support

BYOK = you pay providers directly. Dymo charges only the subscription. Estimated costs in the dashboard use each provider's public rate card.

OpenAI-compatible
14 providers
Cost analytics
Injection + PII guards

Built for developers by developers

World-class docs. Real SDKs. Time-to-first-token measured in seconds.
openai-nodeopenai-pythonai-sdk (Vercel)LangChainLlamaIndex

Frequently Asked Questions

One key. Every model.

Free tier in 60 seconds. Production-ready in five minutes.

Try Dymo Today

Get Started