AMD-Pro-BenchCommunity

13,812 benchmarks in 32 GPU combinations documented

GLM-4.5-Air on AMD GPUs

20 results · 3 sources · 62 measurements

Explore all results

Grouped by reported model identity. Exact weights, quantizations and software can differ. How results are reviewed

Documented results

20 selected tests
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceHumanEval100 problems
63 %Reported pass@1 · n = 100
—
37.33 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQModel ref. ↗Token speed—
466.06 tok/sPrefill · median · unknown
6.17 tok/sGeneration · median · unknown
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Not reported
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
527 / 256 tokens
Context
783
Concurrency
Not reported
Date
2026-07-11
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQ-4bitModel ref. ↗Token speed——
198.31 tok/sGeneration · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
17024 / 8192 tokens
Context
32768
Concurrency
Not reported
Date
2026-07-02
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-bulk-dispatcher-impl——
34.75 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 4096 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQModel ref. ↗Token speed
82.28 tok/sThroughput · peak · aggregate
——1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Not reported
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
34048 / 16384 tokens
Context
32768
Concurrency
64
Date
2026-07-12
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQ-4bitModel ref. ↗Token speed——
341.12 tok/sGeneration · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
34048 / 16384 tokens
Context
32768
Concurrency
Not reported
Date
2026-07-02
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-horner-mpmath-derivative——
35.17 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 3623 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQModel ref. ↗Token speed
103.01 tok/sThroughput · peak · aggregate
——1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Not reported
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
51072 / 24576 tokens
Context
32768
Concurrency
96
Date
2026-07-12
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQ-4bitModel ref. ↗Token speed——
540.87 tok/sGeneration · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
68096 / 32768 tokens
Context
16384
Concurrency
Not reported
Date
2026-07-02
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-invoke-sign-artifacts-ps——
37.64 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 1286 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportLocalMaxxing		GLM-4.5-AirAWQModel ref. ↗Token speed
26.7 tok/sThroughput · peak · aggregate
——1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Not reported
Engine
vllm
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
17024 / 8192 tokens
Context
32768
Concurrency
32
Date
2026-07-12
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-legacy-rename-toolchain——
38.27 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 976 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-nuget-multitarget-design——
38.13 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 1024 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-parser-nunit-tests——
35.55 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 3366 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-poincare-section-pytest——
35.67 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 2872 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-range-expand-bug-fix——
34.84 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 4096 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-relay-command-refactor——
35.77 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 2929 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-resx-toolbar-button——
35.29 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 3485 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-vanderpol-rk4-bugfix——
38.45 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 617 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
View
ReportCustom evaluation		GLM-4.5-AirQ4_K_MModel ref. ↗Conflicting evidenceToken speedsynth-vanderpol-rk4-stepper-c——
34.75 tok/sGeneration · mean · per request
2 × W7800Layer split · 48 GB VRAM/GPU
Configuration
Memory
48 GB VRAM/GPU
OS
Ubuntu 24.04 LTS
Engine
Ollama
Backend
ROCm
Harness
BoJl unified evaluation harness (2026-08)
Prompt / output
? / 4096 tokens
Context
Not reported
Concurrency
Not reported
Date
Not reported
Full record →
Quantizations and software

Quantizations: AWQ, AWQ-4bit, Q4_K_M
Engines: Ollama, vllm

Explore all 20 results →

This is a selection across sources, not a ranking or a matched experiment. Reading benchmark reports

GPU benchmark pages: W7800 · R9700 · Ryzen AI Max+ 395