Skip to content

Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.

Studio: raise the startup bundle budget for #14197 (#14203)

Daniel Han0174c538fb3c
diff --git a/studio/frontend/scripts/check-bundle-budget.ts b/studio/frontend/scripts/check-bundle-budget.ts
index e6683c9d135..6bda08f7b2c 100644
--- a/studio/frontend/scripts/check-bundle-budget.ts
+++ b/studio/frontend/scripts/check-bundle-budget.ts
@@ -244,8 +244,12 @@ export const BUDGET = {
// 2,054,072 -> 2,054,124 B transfer against 4dc8bf0bd2 and 551dc28123, one machine and build, Node 24, chunk
// count unchanged at 95): open-link.ts is the shared link opener the eager shell imports. Raised by that growth,
// again to the measured build plus the same 160 B raw / 189 B transfer margin.
- transferBytes: 2_054_313,
- rawBytes: 6_826_062,
+ // Keeping the FLUX.3 Image Klein GGUF row under its Fast FP8 row (#14197) adds 188 B raw / 66 B transfer
+ // (6,825,902 -> 6,826,090 B raw, 2,054,124 -> 2,054,190 B transfer against 94464babd6 and 1de12ab309, one machine
+ // and build, Node 24, chunk count unchanged at 95): the picker catalog is imported by startup code (chat-adapter.ts).
+ // Raised by that growth, to the measured build plus the same 160 B raw / 189 B transfer margin.
+ transferBytes: 2_054_379,
+ rawBytes: 6_826_250,
};
// The chunk count is reported but not budgeted. Splitting a page out of the entry raises it while lowering the