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.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
Commits on main
Studio passthrough: read the backend's identity through getattr in the cut scope (#13769)
5e927b0 Daniel Han committed 1 month ago
Studio: install the newest audio.cpp release and offer it in updates (#13589)
7b6b2a8 Etherl committed 1 month ago
fix(studio): preserve chat tool selections across model unloads (#13621)
91f8457 Eyera committed 1 month ago
Format studio/backend/routes/inference.py back to the hook's fixed point (#13765)
6694ff8 Daniel Han committed 1 month ago
Studio: re-embed documents after switching the embedding model (#13702)
b9c93fe Long Yixing committed 1 month ago
Studio: deleting a GGUF row no longer deletes other builds at the same quant (#13368)
0b40c81 alkinun committed 1 month ago
Studio: render numbers wrapped in dollar signs as math, not prices (#13649)
525f3ce Nilay committed 1 month ago
Studio: keep the start of an attached file larger than the context window (#13620)
97ed941 Daniel Han committed 1 month ago
Studio: seed the LoRA init with the Random Seed setting (#13637)
9f606d6 Nilay committed 1 month ago
Studio frontend: raise the startup bundle budget by the measured growth since the last raise (#13752)
eba7641 Daniel Han committed 1 month ago
Studio: ask before a desktop Update stops downloads in progress (#13650)
1298fca Nilay committed 1 month ago
Fix Gemma2 sliding-window boundaries across attention backends (#13730)
7453f85 cherry77-cloud committed 1 month ago
Studio: let browser panel pages call their own site's API (#13664)
0cefbf4 oobabooga committed 1 month ago
Studio: give a killed llama-server time to exit before a reload fails (#13622)
c06e4a1 Daniel Han committed 1 month ago
Studio: keep a threadless conversation's cut between requests so the prompt cache is reused (#13743)
06859b3 Daniel Han committed 1 month ago
Fix installer warning colors on light backgrounds (#13572)
5c17163 Lee Jackson committed 1 month ago
Studio: report prompt cache hits on the Messages and Responses APIs, and ask vLLM and SGLang for them (#13741)
ab0abb7 Daniel Han committed 1 month ago
Studio: forget deleted audio models once the Hub cache listing updates (#13630)
dd56fb3 Etherl committed 1 month ago
Studio: move the RAG embedder to CPU when its GPU llama-server keeps crashing (#13739)
1461277 Daniel Han committed 1 month ago
Studio: add an Audio API card to Settings and open it from Audio pages (#12821)
d9214ae Etherl committed 1 month ago
Studio: keep mid-conversation Anthropic system messages in place so new sessions reuse the prompt cache (#13727)
8611f18 Daniel Han committed 1 month ago
Studio: keep Start Export off for a Hub push until the repo is named (#13651)
e4a00b0 Nilay committed 1 month ago
Keep the Docker notebooks and the models they save after docker rm (#13660)
38ca996 Nilay committed 1 month ago
Copy the input in the Gemma inference RMSNorm so float32 decode keeps its residual (#13643)
ae842f7 Vineeth Sai Varikuntla committed 1 month ago
Studio: open the Images and Audio workflows from the collapsed sidebar rail (#13731)
26d6059 Daniel Han committed 1 month ago
Studio: search for llama.cpp and other file names in the address bar (#13661)
1f77626 Nilay committed 1 month ago
Studio: keep a Claude session working after an image on text-only GGUFs (#13652)
7f75c32 Nilay committed 1 month ago
Decode small 4bit batches with the NF4 GEMV instead of dequantizing every token (#13642)
26df950 Daniel Han committed 1 month ago
Studio: say why Sandbox High is unavailable when there is nothing to install (#13732)
047fede Daniel Han committed 1 month ago
Studio: prevent tooltips from hidden run settings (#13614)
021d85e Wasim Yousef Said committed 1 month ago