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
preserve local model inventory after cache filesystem errors
e15867e Lee Jackson committed 2 months ago
fix(desktop): explain updates blocked by an existing studio server
13bdbd9 Lee Jackson committed 2 months ago
studio: tell the model which tool calls were skipped by the per turn limit (#11154)
c98e425 Nilay committed 2 months ago
fetch remote image urls before forwarding to llama-server (#11160)
8b3c658 oobabooga committed 2 months ago
Studio: preserve partial state for canceled GGUF companions (#11018)
4e4ab12 Etherl committed 2 months ago
Studio: preserve Claude Code prompt cache with custom output styles (#11121)
4ec211f Lee Jackson committed 2 months ago
Studio: keep a LoRA push private when the Hub repo already exists (#11229)
c14892b Nilay committed 2 months ago
Route the last direct pwsh spawns through the shared runner (#11303)
bbe4803 Daniel Han committed 2 months ago
Studio: run one repair at a time, however many times Retry is pressed (#11249)
58339d5 Leo Borcherding committed 2 months ago
Route today's new PowerShell test calls through the shared runner (#11285)
e05acf5 Daniel Han committed 2 months ago
studiobench: count the attachments the composer actually renders (#11302)
2c49c88 Daniel Han committed 2 months ago
Assert the companion scan's thread, not how fast the machine is (#11288)
9f76c5a Daniel Han committed 2 months ago
Stop three tests stranding a waiting request in a module-level dict (#11287)
df18fa7 Daniel Han committed 2 months ago
Give the page-fault stall test a window the scheduler cannot close (#11284)
6e262b1 Daniel Han committed 2 months ago
Carry over five temp-scanner fixes that landed after #11217 merged (#11280)
40e9457 Daniel Han committed 2 months ago
Bump install.sh / install.ps1 pin to unsloth>=2026.9.7 (#11279)
ca59c72 Daniel Han committed 2 months ago
Size VRAM from the pool, not the carve-out, on integrated NVIDIA parts (#11208)
77e1a3c Daniel Han committed 2 months ago
Studio: stop the macOS update re-probing binaries and resolving 16 releases (#11237)
df0d65b oobabooga committed 2 months ago
perf(studio): enable the scoped MLX residual normalization and MoE router fusions (#11101)
ef00d49 Long Yixing committed 2 months ago
Studio: stop a Windows update deleting the venv it is running from (#11248)
22192f7 Leo Borcherding committed 2 months ago
Studio: make Save Steps 0 actually turn off checkpoints (#11232)
f250c73 Nilay committed 2 months ago
Make the Colab pip cache able to shrink, which #11270 assumed it already could (#11271)
a6ffbed Daniel Han committed 2 months ago
Studio: probe the GPU once per embedding backend, not on every encode (#11142)
dd25fe1 Wayne Sletcher committed 2 months ago
Contain a temp-scan failure to the directory that raised it, and refuse a partial scan (#11217)
bf61e64 Daniel Han committed 2 months ago