Downloads · 30 days
116
23% of all-time downloads
gemma-4-26B-A4B-it-oQ3e-mtp
mlx-works/gemma-4-26B-A4B-it-oQ3e-mtp
gemma-4-26B-A4B-it-oQ3e-mtp is a machine learning model from mlx-works. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for mlx.
This model was quantized using oQ (oMLX v0.5.7) mixed-precision quantization.
All-time downloads
505
Public
Parameters
26.2B
13.5 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors13.4 GB · 100%
Parameter types
How the weights are stored.
U3225.2B · 96%
Model card
From the Hugging Face model README
gemma-4-26B-A4B-it-oQ3e-mtp
This model was quantized using oQ (oMLX v0.5.7) mixed-precision quantization.
Base model: google/gemma-4-26B-A4B-it
MTP head: google/gemma-4-26B-A4B-it-assistant (merged into single checkpoint)
Quantization details
- Model type: gemma4
- Bits: 3
- Group size: 64
- Format: MLX safetensors
- MTP: Merged from assistant checkpoint (Lightning MTP)
- Calibration: oQ3e (enhanced, imatrix-based; oqe_code_multilingual)
Environment
- Hardware: M5 MacBook Air 32GB
- Inference Framework: oMLX v0.5.7
- Max Concurrent Requests: 4
- Settings:
- Thinking: Disabled
- TurboQuant KV Cache: Enabled (4-bit)
- Lightning MTP: Enabled (key speed improvement)
Performance Benchmarks
Note: Results are for reference only and may vary depending on hardware, software configuration, and workload.
Single Request Results
| Test | TTFT(ms) | TPOT(ms) | pp TPS | tg TPS | E2E(s) | Throughput | Peak Mem |
|---|---|---|---|---|---|---|---|
| pp1024/tg128 | 1065.4 | 23.12 | 961.1 tok/s | 43.6 tok/s | 4.012 | 287.1 tok/s | 11.44 GB |
| pp4096/tg128 | 4695.0 | 24.52 | 872.4 tok/s | 41.1 tok/s | 7.836 | 539.0 tok/s | 11.85 GB |
Continuous Batching (pp1024 / tg128)
| Batch | tg TPS | Speedup | pp TPS | pp TPS/req | TTFT(ms) | E2E(s) |
|---|---|---|---|---|---|---|
| 1x | 43.6 tok/s | 1.00x | 961.1 tok/s | 961.1 tok/s | 1065.4 | 4.012 |
| 2x | 50.7 tok/s | 1.16x | 785.2 tok/s | 392.6 tok/s | 2608.1 | 7.659 |
| 4x | 55.6 tok/s | 1.28x | 676.9 tok/s | 169.2 tok/s | 5856.2 | 15.256 |
Intelligence Benchmark
Note: Each benchmark round tests only 30 questions. Results are for reference only.
| Benchmark | Accuracy | Correct | Total | Time(s) | Think |
|---|---|---|---|---|---|
| MMLU | 63.3% | 19 | 30 | 32 | No |
| TRUTHFULQA | 83.3% | 25 | 30 | 15.2 | No |
| GSM8K | 93.3% | 28 | 30 | 188.9 | No |
| MATHQA | 13.3% | 4 | 30 | 98.5 | No |
| HUMANEVAL | 93.3% | 28 | 30 | 212.4 | No |
- Library
- mlx
- Type
- gemma4
- Created
- Aug 15, 2026
- Updated
- Aug 16, 2026