One AI. The right model for every task.
Within a point of Claude Opus 5.5 on graduate-level science, at a fourteenth of the cost. Built in Berlin.
Free credits on sign-up · Compare plans
Get going with NuRoute in your favourite tool
Point your coding agent, automation or framework at NuRoute, and Artemis picks the model for every call.
- Claude CodeOne line to set up
- piOne line to set up
- n8nCommunity node
- ZapierZapier integration
- MakeMake integration
- LangChainPython package
- CrewAIBuilt-in LLM class
- Vercel AI SDKProvider package
- DifyMarketplace plugin
- MCPMCP server
- JavaScript SDKnpm package
- Python SDKpip package
- Any OpenAI or Anthropic clientChange the base URL
Benchmarks
Artemis, the engine that picks NuRoute’s model, on public tests. Shown at Balanced, the setting NuRoute Chat uses.
- NuRoute, Balanced
- 88.9%
- gpt-5.5
- 92.6%
- NuRoute’s cost
- 2.4%
190 graduate-level science questions. Cost is NuRoute’s as a share of gpt-5.5’s. Hover a point for its numbers.
| Setup | Correct | Per 1,000 |
|---|---|---|
| gemini-3.1-pro | 93.2% | $79 |
| gpt-5.5 | 92.6% | $57 |
| Claude Sonnet 5.5 | 90.5% | $9.31 |
| Artemis, Quality-first | 89.5% | $19 |
| Claude Opus 5.5 | 89.5% | $19 |
| Artemis, Balanced | 88.9% | $1.37 |
| gpt-6-sol | 88.4% | $5.95 |
| Kimi-K2.6 | 88.4% | $82 |
| Artemis, Cost-first | 87.4% | $0.84 |
| gpt-5.6-luna | 87.4% | $1.63 |
| gpt-6-luna | 86.8% | $0.80 |
| Claude Haiku 5.5 | 84.2% | $0.95 |
| Qwen3-235B | 76.3% | $2.80 |
| Nemotron-3.5-Lightning | 76.3% | $3.83 |
| ministral-8b | 53.7% | $0.32 |
- NuRoute, Balanced
- 34.0%
- gpt-5.5
- 40.8%
- NuRoute’s cost
- 3.3%
500 text-only questions. Cost is NuRoute’s as a share of gpt-5.5’s. Hover a point for its numbers.
| Setup | Correct | Per 1,000 |
|---|---|---|
| Claude Opus 5.5 | 56.6% | $64 |
| Artemis, Quality-first | 54.4% | $62 |
| Claude Sonnet 5.5 | 46.4% | $51 |
| gemini-3.1-pro | 45.8% | $202 |
| gpt-6-sol | 41.8% | $17 |
| gpt-5.5 | 40.8% | $139 |
| Kimi-K2.6 | 35.0% | $162 |
| Artemis, Balanced | 34.0% | $4.61 |
| Claude Haiku 5.5 | 31.0% | $3.62 |
| Artemis, Cost-first | 29.6% | $2.46 |
| gpt-6-luna | 28.0% | $1.60 |
| gpt-5.6-luna | 23.8% | $2.82 |
| Qwen3-235B | 12.0% | $3.17 |
| Nemotron-3.5-Lightning | 11.2% | $6.86 |
| ministral-8b | 4.2% | $0.50 |
| Benchmark | NuRoute | Frontier | Cost |
|---|---|---|---|
| RouterArena, vs GPT-5 | 75.5% | 74.0% | 6% |
| GPQA Diamond, vs gpt-5.5 | 88.9% | 92.6% | 2.4% |
| Humanity’s Last Exam, vs gpt-5.5 | 34.0% | 40.8% | 3.3% |
Frontier: the named model answering every question. Cost: NuRoute’s cost as a share of it. RouterArena is independent.
Subscriptions and API access
Chat plans with a weekly token allowance, VAT included, or Artemis on the API, paid per token.
Lite7-day free trial
$3.99a month
- 5M tokens a week
- Chat: 10 requests a minute
- Web search and files
Personal
$10a month
- 50M tokens a week
- API: 60 requests a minute
- Chat: 20 requests a minute
- Web search and files
Personal Max
$100a month
- 600M tokens a week
- API: 120 requests a minute
- Chat: 60 requests a minute
- Web search and files
Artemis API
- /Mtok input
- $0.75
- /Mtok output
- $3.20
- Pay as you go
- One price, whichever model answers
- Artemis pricing
Why we built NuRoute
New models ship every month, and every provider recommends its own.
Nobody’s job is to keep you on the right model for each task. That is NuRoute’s job: for every request, Artemis weighs the models in its pool, across providers, and picks one.
NuRoute for teams in production
The same model selection inside your product, with the controls a company needs.
- Policies
- Allow-lists for models and providers, region rules and PII redaction, checked before a request leaves.
- Budgets
- Monthly budgets with hard limits per project, so spend can’t run away.
- Fallback
- If a provider fails, the next best model and provider takes over, up to three tries.
- Traces
- Every routing decision is recorded, with the candidates considered, and can be replayed.
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.
How NuRoute works
You ask
In your tools, through the API or in chat.
Artemis picks a model
It weighs how well each model would answer this request, and what it would cost.
You get the answer
From the one model that fits the task.
One model answers each request; nothing is blended. Choosing adds no extra AI call. How Artemis decides
Who we are
A Berlin company. We spent four years building machine learning together inside a GDPR-bound health platform, and built NuRoute as the layer we wanted there.
Amit Gupta
CEO
Vincent Bremer
Chief AI Officer
From the blog
Guides
How to Use Multiple AI Providers Through One API
Use multiple LLM providers through one API with NuRoute Enterprise. Learn how this differs from standard Artemis-only access and what to test during migration.
3 min read
Guides
How to Choose an LLM for Each Task Automatically
Build an automatic model-selection workflow: define eligible models, set quality criteria, inspect routing decisions and evaluate real requests.
3 min read
Compare
NuRoute vs Portkey: Model Selection and Gateway Controls
Compare NuRoute and Portkey for model routing, project budgets and gateway operations. See which requirements to test before choosing a platform.
3 min read
Questions
How does NuRoute choose a model?
Artemis, the engine inside NuRoute, reads each request, estimates how well each model in its pool would answer it and what that would cost, and sends it to the best fit. Choosing adds no extra AI call.
Is NuRoute right for every task?
It is built for everyday work: coding, writing, research, analysis and pulling facts out of documents. If you need one specific model for a job, developers can name it through the API.
How should I test NuRoute?
Give it real work you already know the answers to, and compare the results and the cost with the model you use today.
Can I choose the model myself?
Not in NuRoute Chat, where NuRoute does the choosing. Developers using the NuRoute API can name any model in the catalogue instead of letting Artemis pick.
How do the plans work?
Lite is $3.99 a month, Personal $10 and Personal Max $100, VAT included. Each has a weekly token allowance that resets every week; on Personal plans the API draws on it too. Lite starts with a 7-day free trial.
What happens to my prompts?
Our Privacy Policy sets out what is stored, for how long and how it is used. Trust & Security covers hosting, encryption and sub-processors.