Comments
>log in to commentHi everyone, I'm Olivier, the developer behind LowRouter, based in Nantes, France. I wanted my agents to stay on EU-sovereign providers when the data calls for it and to prefer cleaner grids otherwise. I also wanted to know what each run emits, without juggling accounts across providers. You set one base URL and key in OpenCode, OpenClaw, your SDK or your framework, then pick a rule (EU-sovereign providers only, lowest carbon, lowest price) or a specific model. Every response tells you the provider, the region and a carbon estimate. Setup guides for the main agents and frameworks are in the docs. I'd love feedback, especially on the routing rules and the carbon numbers.
LowRouter is a European AI gateway that lets your agents choose where their requests go and shows what each one emits. Each API key follows the rules you set: EU-sovereign providers only, the cleanest electricity grid, the lowest price, or a mix. Every response includes an estimate of its carbon emissions, along with the provider and region that served it.
Point OpenCode, OpenClaw, Claude Code, Cline or Goose at it, or call it from an agent you wrote with the OpenAI SDK, LangChain, CrewAI, the Vercel AI SDK or n8n. You get one key and one bill in euros for language models from many providers, European and international. Aliases let you switch model or provider without changing code or config. In automatic mode, a failed request moves to another provider, so a long run doesn't stop because one provider is down.
Give each agent its own key to see what it spends and emits, and revoke that key without touching the others. The model catalogue is public and free to search by sovereignty and carbon intensity. Tokens are billed at the providers' own prices with no per-token markup; a fee applies when you buy credits, and there is no subscription.
- Carbon‑aware routing with lowest‑carbon provider selection
- EU‑hosted sovereign platform operating from France
- Auto‑routed model IDs (auto/<creator>/<model>)
- Per‑request metadata with energy and carbon estimates
- Unified OpenAI‑compatible API for multiple LLM providers
- Usage monitoring, cost tracking, and consolidated billing
LowRouter is an EU-hosted proxy that routes OpenAI-compatible requests to multiple LLM providers, showing cost and carbon emissions per call.
- for
- Developers building AI agents or apps who need sovereign, carbon-aware routing and unified billing.
- pricing
- paid · free trial
Key features
- Single OpenAI-compatible endpoint — Point any OpenAI-compatible SDK to LowRouter and access 40+ providers with one API key.
- Carbon receipt per request — Each response includes provider, region, gCO₂eq emitted and grid carbon intensity.
- Auto-routing policies — Define rules (e.g., lowest carbon, lowest cost, EU-only) and LowRouter selects the optimal provider.
- Alias management — Create model aliases that can be repointed to different providers or regions without code changes.
- Failover handling — In automatic mode failed requests are rerouted to another provider to keep runs alive.
- Unified EU VAT invoice — All usage across providers is billed on a single euro-denominated invoice.
- Exportable usage reports — Dashboard provides CSV export of tokens, cost and emissions for finance and ESG reporting.
- No prompt retention — LowRouter never stores or trains on your prompts; only metadata for billing is kept.
Use cases
- Route LLM calls from an autonomous agent while staying within EU data-jurisdiction.
- Optimize AI costs and emissions by automatically selecting the greenest provider per request.
- Consolidate billing for multiple LLM providers into one EU-compliant invoice.
- Generate auditable carbon footprints for ESG reporting of AI workloads.
LowRouter vs alternatives
 LowRouter |  n8n |  Mastra |  LlamaIndex |  Postman | |
|---|---|---|---|---|---|
| Best for | EU-sovereign, carbon-transparent LLM routing | AI workflow automation | AI agent framework | Data-aware LLM apps | API testing and design |
| Pricing | Paid | Subscription | Free | Free | Subscription |
| DevHunt upvotes | 1 | 0 | 0 | 0 | 0 |
| Launched | Nov 2026 | — | — | — | — |
- LowRouter vs n8n: n8n is a workflow automation platform; LowRouter focuses on LLM routing with carbon receipts.
- LowRouter vs Mastra: Mastra is a TypeScript framework for building AI agents; LowRouter provides the routing layer for any agent.
- LowRouter vs LlamaIndex: LlamaIndex helps index data for LLM apps, while LowRouter handles provider selection and billing.
- LowRouter vs Postman: Postman is an API design/testing tool; LowRouter is a runtime proxy for LLM calls with sustainability data.
LowRouter FAQ
What is LowRouter?+
LowRouter is a proxy that forwards OpenAI-compatible requests to the LLM provider you choose and returns a carbon and cost receipt.
Do you store or train on my prompts?+
No, prompts are not retained, transformed, or used for training; only billing metadata is kept.
Can I force requests to a specific jurisdiction?+
Yes, you can pin an alias or filter the catalogue to EU-only providers and regions.
How are the carbon numbers calculated?+
They are estimates based on public grid-average data (Ember Climate, regional operators) and the EcoLogits inference model.
What fees apply to usage?+
Tokens are billed at the upstream provider’s per-token price in EUR with no markup; a one-time top-up fee (5.5-8.5% + €0.30) is applied.
Is there a subscription or seat charge?+
No, LowRouter uses a pay-as-you-go model with no monthly commitment or per-seat fees.
Summarized by DevHunt from lowrouter.ai · Nov 2, 2026. Details may change; check the official site.
