Search results

Feature-request: add support VibeThinker

Status: Open.

#1862 In LostRuins/koboldcpp;

· Ricookies707 opened on Nov 27, 2025

Using instruct and mmproj as sdclip1 and sdclip2 loads model twice

Status: Open.

#1861 In LostRuins/koboldcpp;

· tuxArg opened on Nov 27, 2025

High DPI screens confuse initial kobold window

Status: Open.

#1858 In LostRuins/koboldcpp;

· RubenGarcia opened on Nov 23, 2025

turning off kv cache?

Status: Open.

#1848 In LostRuins/koboldcpp;

· blenderman94 opened on Nov 15, 2025

Rope Scaling is missing?

Status: Open.

#1847 In LostRuins/koboldcpp;

· blenderman94 opened on Nov 15, 2025

Embeddings not utilizing GPU

Status: Open.

#1846 In LostRuins/koboldcpp;

· roj234 opened on Nov 15, 2025

Way to control placement of multimodal data in API

Status: Open.

#1845 In LostRuins/koboldcpp;

· Krokozor opened on Nov 14, 2025

[Feature Request] Sageattention over Flashattention

Status: Open.

#1830 In LostRuins/koboldcpp;

· SammCheese opened on Nov 3, 2025

Phi 4 reasoning plus: How to use reasoning?

Status: Open.

#1829 In LostRuins/koboldcpp;

· DoS007 opened on Nov 3, 2025

[Feature Request] Add host-memory prompt caching (Context caching in RAM)

Status: Open.

#1827 In LostRuins/koboldcpp;

· Pento95 opened on Nov 3, 2025

GGML_ASSERT(id >= 0 && id < n_expert) failed when using rowsplit across experts

Status: Open.

#1821 In LostRuins/koboldcpp;

· gpez-git opened on Nov 1, 2025

Koboldcpp 1.100.1 Crashes When Loading Qwen3-Next-80B-A3B-Instruct-GGUF Q8_0 (84.8 GB)

Status: Open.

#1819 In LostRuins/koboldcpp;

· xelercen opened on Oct 30, 2025