https://agent-zone.ai/skills/embedding-pipeline-development/Recent content in Embedding-Pipeline-Development on Agent ZoneHugoen-usWed, 20 May 2026 00:00:00 +0000https://agent-zone.ai/knowledge/serverless/vectorize-id-64-byte-limit-hash-pattern/Wed, 20 May 2026 00:00:00 +0000https://agent-zone.ai/knowledge/serverless/vectorize-id-64-byte-limit-hash-pattern/<h1 id="cloudflare-vectorize-id-64-byte-limit">Cloudflare Vectorize Id 64-Byte Limit<a class="anchor" href="#cloudflare-vectorize-id-64-byte-limit">#</a></h1> <p>Cloudflare Vectorize caps vector ids at <strong>64 BYTES</strong>, not 64 characters. The naive <code>if id.length &lt;= 64</code> skip-hashing check passes Unicode through and then fails at upsert time. The right pattern is unconditional SHA-256 hex hashing with the original id stored in metadata so query results round-trip back to your source-of-truth row.</p> <h2 id="tldr">TL;DR<a class="anchor" href="#tldr">#</a></h2> <ul> <li>The limit is <strong>64 bytes</strong>, not 64 chars. Multibyte UTF-8 hits it sooner than ASCII.</li> <li>Always hash the id. Never branch on length.</li> <li>Put the original id in <code>metadata.id</code>. Resolve back at query time.</li> <li>A single oversized id fails the WHOLE batch — partial-success semantics.</li> </ul> <h2 id="the-error">The error<a class="anchor" href="#the-error">#</a></h2> <pre tabindex="0"><code>VECTOR_UPSERT_ERROR (code = 40008): id too long; max is 64 bytes, got 67 bytes</code></pre><p>This is a 4xx-class refusal at the upsert API. One bad id in a <code>vectorize.upsert([...])</code> batch rejects every vector in the call — it is not partial-success-with-warnings. If you batch 100 vectors and one has a 67-byte id, all 100 silently fail to land.</p>