Both put one API in front of many model providers. The difference is who runs it: GoModel is software you host with your own keys, OpenRouter is a hosted service with one account and one balance. GoModel can also use OpenRouter as one of its providers.
OpenRouter wins on catalog and zero operations. GoModel keeps traffic, keys, and logs in your network.
| GoModel | OpenRouter | |
|---|---|---|
| Who runs it | You, in your network | OpenRouter, as a hosted service |
| How you pay | Providers bill you directly | Credits, 5.5% fee on card purchases |
| Model catalog | 34 providers, your own keys | 500+ models from 80+ providers |
| Self-hosted models | vLLM, SGLang, Ollama, llama.cpp | Not supported |
| Spend control | Budgets and rate limits per user | Per-key limits, org guardrails |
| Caching | Exact and semantic response cache | Prompt caching, exact response cache |
| Operations | You run one container | Nothing to host |
An honest split.
Add OpenRouter as one GoModel provider: heavy traffic goes direct, OpenRouter covers the long tail, and both sit behind the same keys, budgets, and failover.
OPENROUTER_API_KEY=sk-or-...
# Optional: narrow the catalog
OPENROUTER_MODEL_FILTER_INCLUDE=anthropic/*,google/gemini-*,*:freevirtual_models:
- source: claude-sonnet
strategy: failover
targets:
- { model: anthropic/claude-sonnet-4-6 }
- { model: openrouter/anthropic/claude-sonnet-4.6 }Call it as openrouter/anthropic/claude-sonnet-4.6. GoModel sends OpenRouter's app attribution headers for you. OpenRouter figures are from openrouter.ai as of October 2026; it is not in the benchmark because a hosted service has no image to run on the same machine.
Benchmarks are an argument; your traffic is the proof. GoModel speaks the OpenAI and Anthropic APIs natively, so pointing an existing app at it is a base-URL change.
docker run --rm -p 8080:8080 \
-e LOG_FORMAT=text \
-e OPENAI_API_KEY="your-openai-key" \
enterpilot/gomodelfrom openai import OpenAI
client = OpenAI(
base_url="http://localhost:8080/v1",
api_key="your-gomodel-key",
)Full setup, providers, and configuration live in the documentation. Questions? Ask on Discord or book a 30-minute call.