selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

MLExpert

Side-by-side specs, memory needs, licence, runtimes and benchmarks. Add up to 4 models.

--vram
any
--gpus
--quant
--add

On your GPU

  • MLExpert
Property
Overview
TaskChat
Parameters34.7B
Active parameters3.5B
Context length256K
Input → output
Released1 Oct 2026
LicenceApache 2.0
Gated downloadno
Memory and speed
Fits your hardwarepick a GPU above≈22 GB
VRAM at Q4≈22 GB
VRAM at Q8≈37 GB
VRAM at FP16≈69 GB
Est. generation speedpick a GPU above—
Popularity
Downloads / 30 days3.4K
Likes17
Runs with
Jan
KoboldCpp
llama.cpp
LM Studio
LocalAI
Ollama
SGLang
Transformers
vLLM
Ollama tag—
Architecture
ArchitectureQwen3_5MoeForConditionalGeneration
Layers40
Attention / KV heads16 / 2
Vocabulary248 320
Weights dtypebfloat16
GGUF availableyes

Keep comparing

Memory figures are estimates for an 8K context at the chosen quantisation (Q4_K_M for LLMs by default). Speeds are rough, memory-bound estimates. Self-reported benchmark scores come from the model authors.