GLM-5.3-Flash
UD-IQ2_XXS GGUF + mmproj · unsloth/GLM-5.3-Flash-GGUF
queued
| Publisher | unsloth |
|---|---|
| Modality | text+vision LLM |
| Parameters | 320.0B total / 18.0B active 320B total, 18B active (model card). 288 routed experts, 8 per token; 34 linear + 11 sparse-attention layers (config.json). Source: model card. |
| Format | UD-IQ2_XXS GGUF + mmproj |
| Size on disk | 95.93 GiB (103,008,979,940 bytes) |
| Estimated memory to run | ~112.2 GB (estimate; see method) |
| Native context | 1,048,576 tokens (config/card) |
| Engine | llama.cpp · recipe |
| License | mit MIT: commercial OK. |
| Archive folder | official |
| Notes | Tight one-box pick (2-bit). |
Where it runs
| Machine | Verdict | Budget used |
|---|---|---|
| Bosgame M5 (x1) | Tight / offload ≈37.2 tok/s ceiling | ~108-110 GB GPU-usable on Linux with GTT raised (96 GB cap on Windows). Fit budget used here: 108 GB. |
| Bosgame M5 (x2, RPC) | Fits ≈30.5 tok/s ceiling | ~200-225 GB usable across two nodes. Fit budget used here: 210 GB; 210-225 GB = tight. |
| MacBook Pro (M4 Max) | No | macOS caps GPU wired memory by default (~75% = ~27 GB). Raising iogpu.wired_limit_mb to ~32 GB is possible but leaves little for the OS. |
| Mac mini (M4) | No | ~2/3 of 16 GB usable by the GPU by default. Already busy as build box + Hermes/Ollama. |
| omarchy (AMD desktop) | After 128 GB RAM | 7 GB of the 8 GB VRAM for weights. Offload capacity depends on system RAM, which is unknown today; after the 128 GB upgrade ~110 GB of MoE experts could sit in RAM. |
| Ugreen DXP2800 NAS | No | Storage host. Already runs ollama/open-webui containers; only tiny models (e.g. Whisper on CPU) are realistic. |
| MacBook Neo | Unknown | Specs unknown: machine was offline/unreachable on Nov 14 2026. |
Token/s = theoretical ceiling from memory bandwidth and active weight bytes; not a benchmark. Nothing on this page is a benchmark score.