{#if !cacheStats || cacheStats.requests === 0}

No cache data yet. Send a message to a Claude model — prompt-cache usage appears here after the first response.

{:else}

Cache Hit Rate {#if tabTitle} {tabTitle} {/if} {cacheStats.requests} req

Session (this tab) {hitPct}%

{#if cacheStats.last}

Last request {lastHitPct}%

{/if}

Tokens (cumulative)

readCache hits {fmt(cacheStats.cacheReadTokens)}

writeCache writes {fmt(cacheStats.cacheWriteTokens)}

freshUncached input {fmt(uncached)}

Total input {fmt(cacheStats.inputTokens)}

Output {fmt(cacheStats.outputTokens)}

Cache reads cost ~10% of fresh input; writes cost ~25% more. A high hit rate after the first turn means caching is working. Resets on reload.

{/if}