Skip to content

The Open Standard for Generative UI

docs: fix nested <p> hydration errors on Gateway and Benchmarks docs pages (#1420)

devin-ai-integration[bot]a058c65348dd
diff --git a/docs/content/docs/gateway/index.mdx b/docs/content/docs/gateway/index.mdx
index c0ecade6b..92aee8db8 100644
--- a/docs/content/docs/gateway/index.mdx
+++ b/docs/content/docs/gateway/index.mdx
@@ -10,11 +10,11 @@ full: true
<h1 className="m-0 text-5xl font-semibold tracking-tight text-fd-foreground sm:text-5xl">
OpenUI Gateway
</h1>
- <p className="mt-6 text-lg leading-8 text-fd-muted-foreground">
+ <div className="mt-6 text-lg leading-8 text-fd-muted-foreground">
OpenUI Gateway provides a unified API to access models across different providers through a
single endpoint. It adds reliability to OpenUI Lang generation with automatic failover and
model-output correction.
- </p>
+ </div>
<div className="mt-8 flex flex-wrap gap-3">
<a
href="/docs/gateway/generate-openui-lang"
diff --git a/docs/content/docs/openui-lang/benchmarks.mdx b/docs/content/docs/openui-lang/benchmarks.mdx
index b879e0660..76289b350 100644
--- a/docs/content/docs/openui-lang/benchmarks.mdx
+++ b/docs/content/docs/openui-lang/benchmarks.mdx
@@ -132,37 +132,37 @@ The latency advantage compounds with UI complexity. A contact form renders **up
<div className="grid md:grid-cols-2 gap-6 my-6">
<div className="p-5 border border-slate-200 dark:border-slate-800 rounded-xl bg-slate-50 dark:bg-slate-900/40">
<div className="font-semibold mb-2">Model</div>
- <p className="text-sm text-slate-600 dark:text-slate-400">
+ <div className="text-sm text-slate-600 dark:text-slate-400">
GPT-5.2, temperature 0. Same system prompt and user prompt for every scenario. Each format is
derived from the same LLM output, not independently generated.
- </p>
+ </div>
</div>
<div className="p-5 border border-slate-200 dark:border-slate-800 rounded-xl bg-slate-50 dark:bg-slate-900/40">
<div className="font-semibold mb-2">Conversion</div>
- <p className="text-sm text-slate-600 dark:text-slate-400">
+ <div className="text-sm text-slate-600 dark:text-slate-400">
The LLM generates OpenUI Lang. Thesys C1 JSON is a normalized AST projection (`component` +
`props`) that drops parser metadata (`type`, `typeName`, `partial`, `__typename`). The YAML
payload and Vercel JSON-Render output are two serializations of the same json-render spec
projection (`root`, `elements`, optional `state`): JSONL emits RFC 6902 patches, while YAML is
serialized with <code>yaml.stringify(..., &#123; indent: 2 &#125;)</code>.
- </p>
+ </div>
</div>
<div className="p-5 border border-slate-200 dark:border-slate-800 rounded-xl bg-slate-50 dark:bg-slate-900/40">
<div className="font-semibold mb-2">Token Counting</div>
- <p className="text-sm text-slate-600 dark:text-slate-400">
+ <div className="text-sm text-slate-600 dark:text-slate-400">
All formats measured with <code>tiktoken</code> using the <code>gpt-5</code> model encoder —
the same tokenizer family as GPT-5.2. Whitespace and formatting is included as-is in the
count. For YAML, the benchmark counts the document payload only and excludes the outer
<code>yaml-spec</code> fence.
- </p>
+ </div>
</div>
<div className="p-5 border border-slate-200 dark:border-slate-800 rounded-xl bg-slate-50 dark:bg-slate-900/40">
<div className="font-semibold mb-2">Latency Model</div>
- <p className="text-sm text-slate-600 dark:text-slate-400">
+ <div className="text-sm text-slate-600 dark:text-slate-400">
Assumes constant throughput (60 tok/s). Real latency also depends on TTFT and network.
Streaming advantage is most visible for the <em>last element to render</em>, not just overall
time.
- </p>
+ </div>
</div>
</div>