Lexi charges 40% of the savings we generate for you. If STONE lowers the provider cost, the Lexi fee is calculated from that verified saving.

You pay no Lexi fee. If STONE cannot reduce the token count for a particular request, the request passes through at the original provider cost with no additional charge. Zero risk.

Log in to your dashboard and click Add Credits. Payments are processed securely via Stripe. You can add any amount and credits never expire.

Rate limits are based on your lifetime spend and increase automatically as your usage grows.

Yes. BYOK (Bring Your Own Key) is fully supported. Add your provider API key in the dashboard and requests will be billed directly to your provider account. You still get STONE optimization and only pay the Lexi fee on savings.

All data in transit is encrypted with TLS 1.3. Cached data is encrypted at rest with AES-256-GCM and automatically purged based on your retention settings. We never sell or share your data.

Yes. All API keys are hashed with SHA-256 and never stored in plaintext. BYOK provider keys are encrypted at rest and only decrypted in memory during request processing.

Yes. LexiCo AS is a Norwegian company subject to GDPR. We provide data processing agreements, support data export and deletion requests, and maintain full audit trails.

STONE stands for Semantic Token Optimization and Natural Encoding. It is our proprietary engine that restructures your prompts and context to use fewer tokens while preserving the full semantic meaning.

Typical savings are up to 79% depending on the use case. Multi-turn conversations and customer support see the highest savings, while novel research queries see lower but still significant reductions.

Minimal impact. In blind benchmark tests, STONE-optimized requests scored 8.4 out of 9.0 compared to unoptimized requests at 9.0. Most users cannot distinguish between optimized and unoptimized responses.

LexiSaaS supports all major providers: OpenAI (GPT-4o, GPT-4o Mini), Anthropic (Claude 3.5 Sonnet, Haiku), Google (Gemini 2.0 Flash, 1.5 Pro), xAI (Grok), DeepSeek (Chat, Coder), and Meta (Llama). New models are added continuously.

Three steps: 1) Create an account and get your API key. 2) Change your base URL to api.lexisaas.com. 3) Send requests as usual. That's it — no SDK changes, no code refactoring.

Yes. Full SSE (Server-Sent Events) streaming is supported. Set stream: true in your request and receive tokens in real-time, exactly like the native provider API.

Yes. LexiSaaS is a drop-in replacement. Just change the base URL in your OpenAI, Anthropic, or other SDK configuration. Request and response formats are identical to the native provider APIs.