Skip to content
in/guard/out
in/guard/out

Integration

Guardrails & fact-checking across every OpenRouter model

OpenRouter solves model choice – Gemini, Claude, GPT, Llama, DeepSeek, Qwen, and hundreds more behind one API. The proxy solves what that creates: wildly different quality and safety behavior across those models, hitting your users through one integration. Put the proxy in front, and the same guardrail pipeline judges every model’s output by the same standard – whichever lab it came from, whatever changed in last week’s silent model update.

This pairing is where the cost economics get interesting. The reason teams use OpenRouter is to route work to cheaper models; the reason cheaper models feel risky is hallucination and inconsistency. With grounding, fact-checking, and format enforcement on the wire, the quality bar comes from the checks – so the router can chase price while the pipeline holds the line, and the dashboard shows true per-request cost including every guard call.

§01 Setting it up

1. Configure the upstream

Point your proxy key at OpenRouter with your OpenRouter credentials. The proxy forwards requests and attributes token pricing per model.

2. Swap the base URL

Your app calls the proxy with the same OpenAI-compatible SDK it already uses; model selection works as before.

3. Standardize the checks

One policy – PII, grounding, format, safety, agent rules – applies identically whichever model served the request. Model-shopping stops meaning safety-shopping.

4. Compare models on evidence

Per-request violations and per-stage findings, sliced by model, show which cheap models actually hold up on your traffic.

§02 The roster, and what to watch on each

Google Gemini

Gemini 2.5 Pro and Flash – huge context windows that invite huge pastes. Long-context work is where figures drift quietly between page 3 and page 40; deterministic grounding matches every number back to its source.

Anthropic Claude

Opus, Sonnet, and Haiku tiers. Routing Claude through OpenRouter also unlocks the proxy’s full agent stack – tool policy, action grounding, budgets – on Claude-driven agents.

OpenAI GPT

The GPT series and its reasoning variants – often the incumbent your cheaper routes get compared against. Per-model violation stats turn that comparison from vibes into a table.

Meta Llama

The open-weight default. Llama-class models are everywhere from fine-tunes to on-prem serving; the pipeline gives their output the same bar as the frontier tiers.

DeepSeek

V3-class chat and R1-class reasoning at aggressive prices – the classic "cheap but is it safe for customers?" candidate. Exactly the traffic to run through grounding in observe mode first.

Alibaba Qwen

Qwen3 – strong multilingual and coding families. Multilingual traffic deserves note: injection patterns are language-agnostic on the wire, and the classifier stages catch what phrase lists miss.

Mistral

European frontier and open models, a frequent pick where EU data posture matters – pair with PII screening and the GDPR mapping for the full story.

Google Gemma

Google’s open-weight small models – popular for cheap classification and drafting tiers, which is precisely where format enforcement and fact-checks earn their keep.

Moonshot Kimi

K2-class models built for agentic work. Agent-tuned models emit more tool calls per task – more reason for tiers, sequence policies, and per-run budgets.

Zhipu GLM

GLM-4.x – strong coding and agentic performance per dollar. Same story: capable, cheap, and worth watching with real per-request evidence before you enforce.

xAI Grok

Grok’s fast-moving releases change behavior between versions; a constant check standard on the wire means a model swap never silently changes your safety posture.

§03 Frequently asked questions

Which models can I route through the proxy?

Anything OpenRouter serves – Gemini and Gemma, Claude, GPT, Llama, DeepSeek, Qwen, Mistral, Kimi, GLM, Grok, and the long tail. The proxy is model-agnostic: it judges the traffic, not the logo, and per-model pricing metadata keeps cost tracking accurate.

Why put guardrails in front of a router instead of picking safer models?

Because “safer” is not a model property you can verify from a pricing page, and it changes with every model update. Checks on the wire give you a constant standard across a changing model roster – and evidence about which models meet it.

Does cost tracking understand OpenRouter’s per-model pricing?

The proxy prices token usage per model and adds every guard-model sub-call, aggregated per request and per run – the number your budget enforcement reads is the number the dashboard shows.

Can I run different checks for different models?

Checks configure per key, user, and request. A common pattern: stricter grounding on traffic routed to cheaper models, since that is where the risk concentrates.

Does tool calling work through the chain?

Yes – OpenRouter exposes OpenAI-compatible tool calling, so the agent guardrails (tool policy, tiers, grounding, taint, budgets) engage exactly as with OpenAI directly.

§04 Related

Run a cheaper model safely

Route OpenRouter traffic through the proxy so the checks - not the model tier - hold the quality bar. We are running a limited demo - sign up and we will get you in as soon as we can.