Decision Workspace
classeve-rai-infer vs needle-core vs oxillama-quant
Side-by-side comparison of Rust crates
49
classeve-rai-infer
experimentalv0.2.5
CPU inference engine for 4-bit quantized LLMs — hand-written AVX2 GEMM kernels, .raimodel format, speculative decoding
56
needle-core
experimentalv0.3.1
no_std compute kernels for Needle: Cactus-Quants matmul, fast Walsh-Hadamard transform, attention, RoPE
54
oxillama-quant
experimentalv0.1.4
Quantization kernels for all GGUF quantization types
Core Metrics
| classeve-rai-infer | needle-core | oxillama-quant | |
|---|---|---|---|
| Health Score | 49 | 56 | 54 |
| Total Downloads | 81 | 532 | 777 |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 0 | 14 | 34 |
| Releases | 4 | 5 | 5 |
| Last Updated | 31d ago | 20d ago | 53d ago |
| Age | 4m | 4m | 5m |
Health Breakdown
classeve-rai-infer
Maintenance
17
Quality
12
Community
6
Popularity
2
Documentation
12
needle-core
Maintenance
17
Quality
15
Community
9
Popularity
3
Documentation
12
oxillama-quant
Maintenance
15
Quality
14
Community
10
Popularity
3
Documentation
12
Technical Details
| classeve-rai-infer | needle-core | oxillama-quant | |
|---|---|---|---|
| Version | 0.2.5 | 0.3.1 | 0.1.4 |
| Stable (≥1.0) | ✗ No | ✗ No | ✗ No |
| License | Apache-2.0 | MIT | Apache-2.0 |
| Dependencies | 12 | 4 | 7 |
| Crate Size | 522KB | 126KB | 538KB |
| Features | 2 | 5 | 5 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2021 | 2021 | 2021 |
| MSRV | 1.87 | 1.87 | 1.89 |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •needle-core leads with a health score of 56/100, but none of the options score above 80.
- •oxillama-quant has the most downloads (777), suggesting wider adoption.