Benchmark
Submitted byinvocation
Public report
This is the public version of the benchmark. It includes model, runtime, protocol, and throughput information, without telemetry, raw samples, or signing metadata.
Compare throughput at the same token counts. A signed report does not independently verify benchmark execution.
Model
- Canonical ID
- google/gemma-4-26B-A4B-it-qat-q4_0-unquantized
- Identity resolution
- artifact_verified
- Identity verification
- verified
- Verified at
- Name
- Gemma-4 26B-A4B IT (smart Q4_0, QAT-lossless)
- Architecture
- gemma4
- Quantization
- Q4_0
- Artifact
- Provider
- huggingface
- Repo ID
- unsloth/gemma-4-26B-A4B-it-qat-GGUF
- Revision
- 7b92b5b28818151e8669af2e45e88d6086f490dd
- Path
- gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf
- SHA256
- a7c5bc715f5ff8e99a3e8901ce7d2b42b402c669bf24f7c5250747633d0f5891
- Format
- gguf
- Quantization namespace
- gguf
- Quantization scheme
- Q4_0
Runtime
- Name
- llama-cpp
- Version
- b9650 (d8a3f523c)
Benchmark
- Mode
- text
- Chip
- Apple M4 Pro
- Backend
- BLAS,MTL
- Protocol
- Report schema
- computearena-benchmark/1
- Benchmark schema
- computearena-measurements/1
- Throughput schema
- llama-bench-conditioned-pp-tg/1
- Execution layout
- one_process_per_workload
- Decode initial context tokens
- 0
- Conditioning schema
- computearena-conditioning/1
Report metadata
- Schema
- computearena-public-benchmark/1
- Submission ID
- 1fc20a77-9155-43e6-aea8-d94c0943be96
- Run ID
- 20b7443b926b8ac31422a3baaa2c6196
- Benchmark date
- (Unix ms: 1789476084285)
- Submitted by
- invocation