The right model for every task
Artemis picks the model for each request. NuRoute Gateway runs your AI traffic in production.
Three requests, three models
Illustrative“Tag this support ticket”
Small modelSimple task, cheaper model
“Work through a multi-step analysis”
Large modelHard task, stronger model
“Check a customer’s ID form”
EU-hosted modelData must stay in the EU
Your app
One request
model: "auto"
NuRoute Gateway
Step 1
Policy and budget checks
Allowed models, regions, PII redaction and spend limits are applied first.
Step 2 · Artemis by NuRoute
Artemis selects the model
It profiles the task and picks the model with the best expected quality for the cost.
Step 3
Provider and fallback
The request goes to an allowed provider. If it fails, NuRoute tries the next best.
Request trace. Every decision, with the candidates considered, is logged and can be replayed.
Providers
Your keys or NuRoute's
Commercial, European and self-hosted endpoints.
Measured, with the baseline shown
97.5%
lower cost than gpt-5.5, at 97.5% of its quality
Balanced preset, 426 held-out prompts
3.5%
of Claude Sonnet 5’s cost, at the same quality
0.952 against 0.949 mean quality
75.5%
correct on RouterArena, at $0.60 per 1,000 queries
Independent. GPT-5: 74.0% at $10.02
curl https://your-gateway/v1/chat/completions \
-H "Authorization: Bearer aicp-..." \
-d '{ "model": "auto",
"messages": [{ "role": "user", "content": "Summarise this ticket: ..." }] }'