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
Fix three tests red on main after #12431, #12408 and #12382 (#12441)
a334038 Daniel Han committed 1 month ago
Format the desktop contract test to the ruff-format fixed point (#12440)
8492a02 Daniel Han committed 1 month ago
fix(studio): serve desktop SPA on loopback listener (#10847)
9dd4977 Sourav Rajvi committed 1 month ago
Count the final EOS token within the raw-text chunk budget (#11040)
a809a51 Vineeth Sai Varikuntla committed 1 month ago
Bump install.sh / install.ps1 pin to unsloth>=2026.9.14 (#12439)
407246c Daniel Han committed 1 month ago
Studio: store the full build and workflow in each image recipe and show them in the Recipe popover (#12091)
16c1129 Sourav Rajvi committed 1 month ago
Resolve a sentence-transformers modules.json module class through the same trust gate as upstream (#12436)
d0b3a7b Daniel Han committed 1 month ago
Collapse the llama.cpp install branch that never branched (#10268)
1efc294 Vineeth Sai Varikuntla committed 1 month ago
Studio: retry custom gateways with max_completion_tokens after a max_tokens 400 (#10837)
5c239a2 Sourav Rajvi committed 1 month ago
Prefetch Tauri NSIS and WebView2 tools before the Windows desktop build (#12434)
1b4981d Daniel Han committed 1 month ago
Keep the GRPO eval batch a multiple of num_generations (#12362)
1f65728 Vineeth Sai Varikuntla committed 1 month ago
Size the packed attention mask at the padded length, not the token count (#8278)
9cf2318 Vineeth Sai Varikuntla committed 1 month ago
Make the registry's default quant_type usable (#9887)
11de7e7 Vineeth Sai Varikuntla committed 1 month ago
Studio: browse temporary Linux mounts under /media and /mnt (#10048)
42b7de3 Sourav Rajvi committed 1 month ago
Studio: keep rounded boxes rounded when they scroll (#12431)
9cf98aa Michael Han committed 1 month ago
Studio: use the resolved public id in embeddings/completions API monitor rows (#9346)
acc0c89 Sourav Rajvi committed 1 month ago
Bump install.sh / install.ps1 pins to unsloth>=2026.9.13, unsloth-zoo>=2026.9.9 (#12432)
7474279 Daniel Han committed 1 month ago
perf(studio): reuse embeddings for identical files in linked folders (#11972)
7bfd700 Tom Mathews committed 1 month ago
Studio: keep PyTorch mirror leaves out of query tokens (#10546)
150e6df alkinun committed 1 month ago
Unsloth Studio: return the spoken text from /audio/generate instead of a truncated label (#12386)
abf4cb4 Leo Borcherding committed 1 month ago
feat(studio): configurable RAG upload extensions via RAG_UPLOAD_EXTS (#11499)
38ef21f Sourav Rajvi committed 1 month ago
Read the desktop New chat button contract token by token (#12426)
813cee7 Daniel Han committed 1 month ago
Studio: Qwen-Image-2.1 placement from measured sizes, int8 under offload on torchao 0.17, balanced fit check (#12408)
cf89ed7 Daniel Han committed 1 month ago
Studio: explain the all-columns-dropped recipe error in UI terms (#12416)
37d6154 Daniel Han committed 1 month ago
Studio: move Blender MCP setup out of Manage MCP servers into the composer (#12430)
a768954 Nilay committed 1 month ago
Studio: keep the 1024 canvas when auto precision picked the Qwen-Image-2.1 quant (#12428)
23adbb2 Daniel Han committed 1 month ago