Audit11 September 2026·5 min

One real invoice: $2.2k a month of LLM, and what compression gets back

An Azure bill from a B2B SaaS, read line by line. A 10 % cache-hit rate, one LLM call per candidate, a timestamp at the top of the system prompt — and an honest estimate of what a compressor alone recovers.

10 %cache-hit rate found

A B2B SaaS shared its February 2026 Azure invoice and read access to its codebase. Foundry Models: $2,218. Of that, uncached input $1,140 for about 912M tokens, output $683 for 68M, and cached input $13 for 106M tokens. That last line is the one to read first: a cache-hit rate around 10 %, down from 21 % in January. You can read it straight off the invoice, no application log needed.

Where the tokens come from

  • Candidate matching: one LLM call per candidate, 52 per search, in parallel, on a model without caching.
  • Document adaptation: one call per work experience, 43 per session.
  • Document generation: 13 parallel calls, each with a variable system prompt placed before the identical résumé — so no shared prefix.
  • Document analysis: one call per page on the largest model.
  • Summaries with reasoning effort set to high.
  • Chat: the date to the minute, memories and tool context at the top of the system prompt, ahead of 10k tokens of stable instructions — the cache only serves steps of the same turn.

The application's own usage table captured about 20 % of the volume and priced one model 13× too high, so the admin dashboard attributed 70 % of spend to matching. It is closer to 17 %.

What each lever is worth

LeverEstimated saving / monthQuality risk
Cacheable prefix (reorder system prompt, freeze timestamps)$400–600none
Batching fan-out calls$100–200none
Output and reasoning settings$130–270low
Structural JSON compaction (lossless)$100–135none
Neural text compression at −20 %$20–50low
Prose is about 10 % of this product's input tokens; the rest is JSON, résumés and tool payloads.

On this invoice a compressor alone is worth $20–50 a month. The cache and the fan-out are worth $600–1,000. That is why the Tokeen SDK ships prefix ordering, batching and JSON compaction next to the compressor — and why the first thing we do is read your invoice.

The five-minute qualification

Two lines on any provider invoice — cached input and uncached input — give the hit rate. Below 30 % on a multi-turn product, the prefix is unstable and the fix is structural. Above 80 %, compression on cold prefixes is the remaining lever. Send us yours; we answer with a number.

More research

Stop paying for filler. Compress every prompt on one fast, deterministic layer.

Tokeen is in private beta. Request access and see the savings on your own prompts within the week.