Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
| 1 model | microsoftmicrosoft/Phi-4-multimodal-instructRemove × |
|---|---|
| What’s different | 6B parameters · 8 languages · Encoder: Conformer-based · Decoder: Phi-4-Mini-Instruct |
| Date | Oct 17, 2026Discovered |
API prices USD · provider configuration shown below | |
| Listed rate | See provider quotes |
| Provider endpoints | 0 endpoints → |
Benchmarks Each result retains its evaluation configuration | |
| AMI-Cleaned WER ↗English short-form; normalization unversioned | 9.12% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| Earnings22-Cleaned-AA-chunked WER ↗English short-form; normalization unversioned | 7.09% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| Gigaspeech-Cleaned WER ↗English short-form; normalization unversioned | 7.79% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| LS Clean WER ↗English short-form; normalization unversioned | 1.35% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| LS Other WER ↗English short-form; normalization unversioned | 3.45% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| SPGISpeech WER ↗English short-form; normalization unversioned | 2.73% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| Voice Arena Monsoon WER ↗English short-form; normalization unversioned | 4.46% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
| Voxpopuli-AA-Cleaned WER ↗English short-form; normalization unversioned | 4.18% WER ↗microsoft/Phi-4-multimodal-instructLeaderboard entry |
Specifications | |
| Context window | — |
| Max output | — |
| Parameters | 6B |
| Audio input | Supported |
| Transcription | Supported |