Phi-5
microsoft/phi-5by Microsoft·Text·128K context·Released 2026-02-09·MIT
Overview
Microsoft's small open-weight model. Trained heavily on curated + synthetic data, so it punches above its size on reasoning but knows less about the world. Very cheap to run. Good for tightly-scoped work: classification, routing, structured extraction.
Benchmarks
Standardized scores (0–100), normalized across the catalog.
Intelligence52
Coding64
Instruction following85
Providers
LLMRouteHub routes to the cheapest healthy provider by default.
| Provider | Input | Output | Throughput | Uptime |
|---|---|---|---|---|
| DeepInfra | $0.08 | $0.24 | 151 tok/s | 99.6% |
| Together | $0.10 | $0.29 | 140 tok/s | 99.4% |
| Fireworks | $0.12 | $0.35 | 130 tok/s | 99.3% |
| Novita | $0.12 | $0.37 | 119 tok/s | 99.1% |
| Groq | $0.14 | $0.42 | 109 tok/s | 99% |
Use this model
curl https://llmroutehub.com/api/v1/chat/completions \
-H "Authorization: Bearer $LLMROUTEHUB_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "microsoft/phi-5",
"messages": [{ "role": "user", "content": "Hello" }]
}'