Transparent pricing. Enterprise-grade value.
Lume is priced per deployment, not per seat. You bring your own models and cloud — we don't charge for data volume or API calls. Every tier includes full RBAC, AD integration, and version-controlled deployments.
You control model costs — we don't mark them up
Lume is model-agnostic. Connect to OpenAI, Anthropic, Azure OpenAI, AWS Bedrock, or self-hosted models. You pay your cloud provider directly — Lume adds no markup on token usage.
Bring your own API keys
Connect to any model provider using your own credentials
Pay your cloud, not us
Model inference costs go directly to your cloud bill
Context window control
Lume manages prompt context condensation to keep costs predictable
Context condensation: Lume automatically summarises long conversation histories to reduce token consumption, keeping AI interactions cost-efficient at scale.
Enterprise security included at every tier
Frequently asked questions
- Do I need to move my data to Lume's cloud?
- No. Lume deploys inside your existing AWS or Azure environment. Your data never leaves your cloud tenancy.
- Which LLMs can I use?
- Any model accessible via API — OpenAI, Anthropic Claude, Azure OpenAI, AWS Bedrock, Google Vertex, or self-hosted open-source models.
- How are model costs calculated?
- You connect your own API keys. Token usage is billed directly by your model provider. Lume adds no markup.
- Can I start with one department and expand?
- Yes. Lume is structured around workspaces and access groups, so you can pilot with one team and roll out progressively.
- What does 'human-in-the-loop' mean for dev agents?
- You configure review checkpoints — for example, requiring a human to approve a PR before it merges. Agents work autonomously up to the review gate.
Ready to see the numbers?
Book a personalised walkthrough and we'll put together a quote based on your environment.