Compare · ModelsLive · 2 picked · head to head
GLM 4 32B vs Hermes 2 Pro - Llama-3 8B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GLM 4 32B wins on 4/6 benchmarks
GLM 4 32B wins 4 of 6 shared benchmarks. Leads in general · knowledge · reasoning.
Category leads
general·GLM 4 32B knowledge·GLM 4 32B language·Hermes 2 Pro - Llama-3 8Bmath·Hermes 2 Pro - Llama-3 8Breasoning·GLM 4 32B
Hype vs Reality
Attention vs performance
GLM 4 32B
#218 by perf·no signal
Hermes 2 Pro - Llama-3 8B
#211 by perf·no signal
Best value
GLM 4 32B
1.1x better value than Hermes 2 Pro - Llama-3 8B
GLM 4 32B
180.0 pts/$
$0.10/M
Hermes 2 Pro - Llama-3 8B
157.9 pts/$
$0.14/M
Vendor risk
Who is behind the model
z-ai
private · undisclosed
nousresearch
private · undisclosed
Head to head
6 benchmarks · 2 models
GLM 4 32B Hermes 2 Pro - Llama-3 8B
BBH (HuggingFace)
GLM 4 32B leads by +5.1
GLM 4 32B
35.8
Hermes 2 Pro - Llama-3 8B
30.7
GPQA
GLM 4 32B leads by +3.1
GLM 4 32B
8.8
Hermes 2 Pro - Llama-3 8B
5.7
IFEval
Hermes 2 Pro - Llama-3 8B leads by +39.4
GLM 4 32B
14.3
Hermes 2 Pro - Llama-3 8B
53.6
MATH Level 5
Hermes 2 Pro - Llama-3 8B leads by +8.4
GLM 4 32B
0.0
Hermes 2 Pro - Llama-3 8B
8.4
MMLU-PRO
GLM 4 32B leads by +12.1
GLM 4 32B
34.9
Hermes 2 Pro - Llama-3 8B
22.8
MUSR
GLM 4 32B leads by +2.9
GLM 4 32B
14.2
Hermes 2 Pro - Llama-3 8B
11.3
Full benchmark table
| Benchmark | GLM 4 32B | Hermes 2 Pro - Llama-3 8B |
|---|---|---|
BBH (HuggingFace) | 35.8 | 30.7 |
GPQA | 8.8 | 5.7 |
IFEval | 14.3 | 53.6 |
MATH Level 5 | 0.0 | 8.4 |
MMLU-PRO | 34.9 | 22.8 |
MUSR | 14.2 | 11.3 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.10 | $0.10 | 128K tokens (~64 books) | $1.00 | |
| $0.14 | $0.14 | 8K tokens (~4 books) | $1.40 |