summaryrefslogtreecommitdiffhomepage
path: root/packages/api/src/routes/models.ts
diff options
context:
space:
mode:
authorAdam Malczewski <[email protected]>2026-06-02 22:50:11 +0900
committerAdam Malczewski <[email protected]>2026-06-02 22:50:11 +0900
commit66e5d3b105bfd2b34c6f35876bf33dbb3cb9dcae (patch)
treec3e039e09c89231f84dfd16f7bbbf8aedcc2dc7d /packages/api/src/routes/models.ts
parent4b45d33c256cf580a53054078be6fd7148fa6302 (diff)
downloaddispatch-66e5d3b105bfd2b34c6f35876bf33dbb3cb9dcae.tar.gz
dispatch-66e5d3b105bfd2b34c6f35876bf33dbb3cb9dcae.zip
feat(chat): paste-to-attach images/PDFs with model capability check
Add multimodal image/PDF input to the chat box via clipboard paste, gated by a graceful per-model capability check. UX: a pasted image/PDF inserts an inline token (【image:…】 / 【pdf:…】) into the draft, so attachments have ORDER relative to typed text and can be referenced positionally. The token is the only handle — deleting it (atomic Backspace/ Delete, or selection overlap) detaches the file; an input-reconciliation safety net detaches any attachment whose token is no longer intact. No preview strip. Capability check: resolveModelCapabilities reads models.dev modalities.input (new GET /models/capabilities, mirrors /context-limit). The input blocks Send (no tokens spent) only on a definitive 'no'; unknown capability (catalog offline / unmapped provider) stays permissive. Attachments require a fresh turn — Send is blocked while generating and /chat rejects content mid-turn (409). Attachments are EPHEMERAL: forwarded to the model for the turn via ordered AI SDK ImagePart/FilePart content, but never persisted (history keeps the text with [image]/[pdf] markers). Text-only turns serialize byte-identically to before. Limits (Anthropic-aligned, enforced at paste + re-validated server-side): PNG/JPEG/WebP/GIF/PDF; image ≤5MB, PDF ≤32MB, ≤20 attachments, ≤32MB total. core: UserContentPart types, models/attachments validator, capability resolver, agent.run+toModelMessages thread ordered content. api: /chat content validation + passthrough. frontend: attachment-tokens helper, ChatInput paste/token/gating, per-tab staged attachments, App.svelte capability fetch. +44 tests.
Diffstat (limited to 'packages/api/src/routes/models.ts')
-rw-r--r--packages/api/src/routes/models.ts18
1 files changed, 18 insertions, 0 deletions
diff --git a/packages/api/src/routes/models.ts b/packages/api/src/routes/models.ts
index eeb6029..a1700b1 100644
--- a/packages/api/src/routes/models.ts
+++ b/packages/api/src/routes/models.ts
@@ -20,6 +20,7 @@ import {
refreshAccountCredentialsAsync,
resolveApiKey,
resolveContextLimit,
+ resolveModelCapabilities,
selectHaikuModel,
setApiKey,
validateAccountCredentials,
@@ -180,6 +181,23 @@ modelsRoutes.get("/context-limit", async (c) => {
return c.json({ contextLimit });
});
+// Resolve a model's image / PDF INPUT capabilities from the models.dev catalog.
+// Returns `{ capabilities: { image, pdf } | null }`. `null` means UNKNOWN — the
+// provider is unmapped, the model is absent, the catalog predates the
+// `modalities` field, or the catalog is offline. The frontend treats `null` as
+// "can't verify" (optimistic allow) and a definitive `{ image: false }` as a
+// hard block (no tokens spent).
+modelsRoutes.get("/capabilities", async (c) => {
+ const provider = c.req.query("provider");
+ const modelId = c.req.query("modelId");
+ if (!provider || !modelId) {
+ return c.json({ error: "provider and modelId query parameters are required" }, 400);
+ }
+
+ const capabilities = await resolveModelCapabilities(provider, modelId);
+ return c.json({ capabilities });
+});
+
// List available Claude accounts with validated credentials
modelsRoutes.get("/claude-accounts", async (c) => {
const candidates = resolveClaudeAccounts();