GLM-4.7-Flash with Hermes Agent

Q4_K_M (GGUF) on llama.cpp commit a37cc46. Tested

Task completion

93.3% (28/30)

Tool-call success

66.7%

Decode tok/s

63.4

Prefill tok/s

2649.6

Server memory

20.5 GB

Avg turns

5.3

Setup

Model: GLM-4.7-Flash
Family / maker: GLM · Z.ai
Parameters: To be confirmed
Architecture: MoE
Weights: unsloth/GLM-4.7-Flash-GGUF
Quant: Q4_K_M (GGUF)
Engine: llama.cpp commit a37cc46
Hermes Agent: v0.22.4
Tested: 2026-10-23
Context: 64K (65,536 tokens); long-context task t09 at 128K or 192K
Memory tier: ≤24 GB (measured)

Parameter count and memory tier to be confirmed from the model card.

Per-task results

t01: 3/3
t02: 3/3
t03: 3/3
t04: 3/3
t05: 3/3
t06: 3/3
t07: 3/3
t08: 2/3
t09: 2/3
t10: 3/3

Copy-paste config

1. Start the model server (llama.cpp)

Exact command used in our test. Download the GGUF from unsloth/GLM-4.7-Flash-GGUF into ./models/.

llama-server --model ./models/GLM-4.7-Flash-Q4_K_M.gguf --alias glm-4-7-flash --host 127.0.0.1 --port 18080 --ctx-size 196608 --parallel 1 --n-gpu-layers 999 --flash-attn on --jinja

2. Point Hermes Agent at it

hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:18080/v1
hermes config set model.default glm-4-7-flash