Last updated: 2026-09-06.
SaveToken ("we," "us," the "Service") is an API that reduces the number of tokens in text you send it — JSON arrays, tool/MCP schemas, and logs/tracebacks — using deterministic, programmatically-verified transforms (structural re-encoding, schema compaction, traceback frame-filtering). We do not call any AI model ourselves. We return the reduced text to you; you send it to your own LLM provider (Anthropic, OpenAI, or otherwise) using your own account and key.
You're responsible for keeping your API key confidential and for all activity under it. Free-tier keys are rate-limited (100 calls/day); paid keys are limited to 5,000 calls/month; all keys carry a burst limit (300 calls/minute). We may suspend or revoke a key without notice for abuse, key-farming, or attempts to circumvent quotas.
Paid plans are billed in advance on the cycle you choose at checkout — monthly, quarterly, or annual — via Paddle, our payment processor and merchant of record, which handles payment collection, tax, and invoicing. We may change quota limits or pricing for future billing periods with notice. Cancel anytime — no lock-in.
Cancelling stops future billing but does not automatically refund the current billing period. If you believe you were charged in error, or the Service was unavailable or materially not as described for a meaningful part of your billing period, contact contact@savetoken.org within 7 days of the charge — we review these individually and refund at our discretion. Refunds, when issued, are processed by Paddle back to your original payment method.
We process the text you submit only to produce the optimized response. We do not store the content of your requests or responses. We log token counts, dollars saved, verdict, and mode per call for your usage dashboard — never the text itself. We never see or store your downstream LLM provider's API key.
SaveToken is an independent product and is not affiliated with, endorsed by, or sponsored by Anthropic, OpenAI, or any other AI provider named on this site or in our documentation.
The Service is provided "as is." We do not guarantee that every request will be reduced in size — content that can't be safely reduced is passed through unchanged (verdict: NEUTRAL). We do not warrant uninterrupted or error-free operation.
To the maximum extent permitted by law, we are not liable for lost profits, broken workflows, downtime, or other consequential damages arising from use or inability to use the Service. Our aggregate liability for any claim is capped at the fees you actually paid us in the 30 days preceding the claim.
You agree to indemnify us against third-party claims arising from how you configure your pipelines, the data you send us, or your use of the optimized output.
Either party may terminate at any time. We may suspend or terminate access for breach of these terms, including abuse of usage limits.
We may update these terms; continued use of the Service after an update constitutes acceptance.
These terms are governed by the laws of India. Courts in Gurugram, Haryana have exclusive jurisdiction over any dispute arising from these terms.
SaveToken, c/o Invictus International Consulting Services, Gurugram, Haryana, India. Contact: contact@savetoken.org