Microsoft

Models
36
Hosted by a direct provider
27

Provider counts exclude gateways.

ModelCompareTypeReleasedContextProvidersInput $/1MOutput $/1M
Microsoft-Decision-1Decisions2026-10-0932.8K0 +1n/a
MAI-Transcribe-2-StreamingTranscription2026-10-010 +1n/a
MAI-Voice-2.1Speech synthesis2026-10-010 +2n/a
MAI-Voice-2.1-FlashSpeech synthesis2026-10-010 +2n/a
MAI-Image-2.6Image generation2026-09-044.1K1 +2n/a
MAI-Image-2.6-FlashImage generation2026-09-044.1K1 +2n/a
MAI-Transcribe 2Transcription2026-09-031 +2n/a
MAI-Code-1.1-FlashChat2026-08-11256K0 +1via GitHub Copilot $0.20via GitHub Copilot $1.20
MAI-Image-2.5 ProImage generation2026-07-2332K1 +1n/a
MAI-Voice-2-FlashSpeech synthesis2026-07-230 +2n/a
MAI-Code-1-FlashChat2026-06-02256K0 +1via GitHub Copilot $0.75via GitHub Copilot $4.50
MAI-Image-2.5Image generation2026-06-0232K1 +1n/a
MAI-Transcribe 1.5Transcription2026-06-020 +1n/a
MAI-Voice-2Speech synthesis2026-06-020 +2n/a
Phi 4 MultimodalChat2025-07-26131K2 +1$0.08$0.32
Phi-4Chat2025-01-1016.4K3 +2$0.070–$0.125$0.14–$0.50
Phi-4-miniunverifiedChat2024-12-11128K2$0.075$0.30
Phi-4-mini-reasoningunverifiedChat2024-12-11128K2$0.075$0.30
Phi-4-multimodalunverifiedChat2024-12-11128K2$0.08$0.32
Phi-4-reasoningunverifiedChat2024-12-1132K2$0.125$0.50
Phi-4-reasoning-plusunverifiedChat2024-12-1132K2$0.125$0.50
Phi-4-MiniChat2024-12-01131K2 +1$0.075$0.30
WizardLM-2 8x22BChat2024-04-1565.5K3 +2$0.48–$1.20$0.48–$1.20
phi-2Other2023-12-121$0.10$0.10
phi-2-3bunverifiedChat—2K1$0.10$0.10
Phi-3.5 Mini 128K InstructunverifiedOther—1$0.10$0.10
phi-3-5-mini-instructunverifiedChat—128K1
phi-3-5-moe-instructunverifiedChat—128K1
phi-3-5-vision-instructunverifiedChat—128K1
Phi-3 Medium 128K InstructunverifiedChat—128K1
phi-3-medium-4k-instructunverifiedChat—4.1K1
Phi-3 Mini 128K InstructunverifiedChat—131K2$0.10$0.10
phi-3-mini-4k-instructunverifiedChat—4.1K1
phi-3-small-128k-instructunverifiedChat—128K1
phi-3-small-8k-instructunverifiedChat—8.2K1
phi-3-vision-128k-instructunverifiedChat—32.1K1$0.20$0.20

Providers: direct count, then “+n” gateways. Prices are the range across comparable direct providers (gateways, subscription plans, deprecated endpoints and anything that is not a paid per-token price excluded); models only on gateways show the cheapest gateway list price. Non-token models show their native unit.