luulam/9router
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(models): expose snake_case token limits on /v1/models

Browse Source
This commit is contained in:
haumanto
2026-08-13 12:18:38 +07:00
committed by decolua
parent 67271d859e
commit 30fec4318e
1 changed files with 21 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
src/app/api/v1/models/route.js
View File

@@ -485,6 +485,27 @@ export async function buildModelsList(kindFilter, options = {}) {
|| capabilitiesFromServiceKind(customKind || liveKind)
|| (kind === LLM_KIND ? getCapabilitiesForModel(providerId, modelId) : null);
if (caps) model.capabilities = caps;
// Token limits under the snake_case names the OpenAI/OpenRouter
// convention uses. `capabilities.contextWindow` is camelCase and nested,
// so clients matching context_length find nothing, fall back to guessing
// the window from the model name, and guess high — a 372k model read as
// 1.05M never reaches its compaction threshold and hard-fails upstream.
// Emitted at top level because not every client recurses into nested
// objects; the camelCase `capabilities` block stays for compatibility.
if (kind === LLM_KIND || allowAsLlm) {
let contextWindow = caps?.contextWindow;
let maxOutput = caps?.maxOutput;
// Live-catalog and service-kind capabilities are usually partial
// (often just { tools: true }), so fill the gaps from the static
// table rather than emitting null and leaving clients to guess.
if (!Number.isFinite(contextWindow) || !Number.isFinite(maxOutput)) {
const fallback = getCapabilitiesForModel(providerId, modelId);
if (!Number.isFinite(contextWindow)) contextWindow = fallback.contextWindow;
if (!Number.isFinite(maxOutput)) maxOutput = fallback.maxOutput;
}
if (Number.isFinite(contextWindow)) model.context_length = contextWindow;
if (Number.isFinite(maxOutput)) model.max_completion_tokens = maxOutput;
}
models.push(model);
}