Decision Workspace
llm-bench vs umf vs dynamo-protocols
Side-by-side comparison of Rust crates
52
llm-bench
experimentalv0.3.0
Send the same prompts to OpenAI, Anthropic, Ollama, OpenRouter, Gemini or any OpenAI-compatible server and compare time to first token, latency percentiles, tokens per second and cost; CLI and library
66
umf
experimentalv0.3.0
Universal Message Format (UMF) - Provider-agnostic message representation for LLM interactions with ChatML formatting, internal hub model, and MCP support
65
dynamo-protocols
growingv8.1.0
Protocol types for OpenAI-compatible inference APIs with inference-serving extensions.
Core Metrics
| llm-bench | umf | dynamo-protocols | |
|---|---|---|---|
| Health Score | 52 | 66 | 65 |
| Total Downloads | 78 | 2.3K | 763.8K |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 0 | 609 | 112 |
| Releases | 6 | 14 | 30 |
| Last Updated | 6d ago | 35d ago | 1d ago |
| Age | 14d | 11m | 6m |
Health Breakdown
llm-bench
Maintenance
16
Quality
13
Community
6
Popularity
2
Documentation
15
umf
Maintenance
17
Quality
16
Community
14
Popularity
4
Documentation
15
dynamo-protocols
Maintenance
19
Quality
16
Community
13
Popularity
7
Documentation
10
Technical Details
| llm-bench | umf | dynamo-protocols | |
|---|---|---|---|
| Version | 0.3.0 | 0.3.0 | 8.1.0 |
| Stable (≥1.0) | ✗ No | ✗ No | ✓ Yes |
| License | MIT | MIT OR Apache-2.0 | Apache-2.0 AND MIT |
| Dependencies | 19 | 6 | 14 |
| Crate Size | 68KB | 39KB | 70KB |
| Features | 2 | 7 | 1 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2021 | 2021 | 2024 |
| MSRV | 1.87 | 1.70 | — |
| Owners | 1 | 1 | 4 |
Links
Quick Verdict
- •umf leads with a health score of 66/100, but none of the options score above 80.
- •dynamo-protocols has the most downloads (763.8K), suggesting wider adoption.
- •umf is depended on by 609 crates — strongest ecosystem trust.
- •llm-bench, umf are pre-1.0 — API may change.