https://agent-zone.ai/tools/anthropic/Recent content in Anthropic on Agent ZoneHugoen-usWed, 20 May 2026 00:00:00 +0000https://agent-zone.ai/knowledge/agent-tooling/cost-per-pass-not-cost-per-call/Wed, 20 May 2026 00:00:00 +0000https://agent-zone.ai/knowledge/agent-tooling/cost-per-pass-not-cost-per-call/<h1 id="cost-per-pass-not-cost-per-call">Cost-Per-Pass, Not Cost-Per-Call<a class="anchor" href="#cost-per-pass-not-cost-per-call">#</a></h1> <p>Practitioners price LLMs by the per-token rate on the provider&rsquo;s pricing page. For autonomous agents, that number is misleading. Two layers of indirection sit between the per-token rate and the cost you actually pay to get work done: variable prompt sizes turn per-token into per-call, and variable pass rates turn per-call into per-pass. Each layer can invert the ranking.</p> <p>For autonomous fleets where failed attempts trigger reviewer cycles, retries, and reputational drag, <strong>cost-per-pass is the only metric that ranks models correctly</strong>. This article shows how to compute it, when it dominates, and where the cheapest-per-token model becomes the most expensive in production.</p>