Model vendor
Technology company
Updated Aug 10, 2026. Data as of 2026-08-07
Default-version specifications and scored-version capability values are labeled separately.
| MI | MAI-Thinking-1 | 63.4 | 80.0% | N/A | Closed weights | MAI-Thinking-1 | |
| MI | Phi 4 Reasoning Plus | 50.3 | 60.0% | N/A | Closed weights | Phi 4 Reasoning Plus | |
| MI | Phi 4 Reasoning | 47.9 | 60.0% | N/A | Closed weights | Phi 4 Reasoning | |
| MI | Phi-3.5-MoE-instruct | 42.4 | 80.0% | N/A | Closed weights | Phi-3.5-MoE-instruct | |
| MI | Phi 4 | 36.9 | 80.0% | 16K | Closed weights | Phi 4 | |
| MI | Phi 4 Mini | 31.5 | 60.0% | 128K | Open weights | Phi 4 Mini | |
| MI | Phi-3.5-mini-instruct | 25.9 | 80.0% | 128K | Closed weights | Phi-3.5-mini-instruct | |
| MI | MAI-Code-1-Flash | N/A | N/A | 256K | Closed weights | Not scored | |
| MI | Phi-3.5-vision-instruct | N/A | N/A | N/A | Closed weights | Not scored | |
| MI | Phi 4 Mini Reasoning | N/A | N/A | N/A | Closed weights | Not scored | |
| MI | Phi-4-multimodal-instruct | N/A | N/A | 128K | Closed weights | Not scored |
Each unique model uses its latest scored version.
Leaders from this vendor's current published catalog.
Vendor context from the published model registry.
Technology company
Model ownership is separate from API hosting. One model can be offered by multiple API providers with different prices.
Vendor description and inventory come from the published model registry.
Common questions about Microsoft.
MAI Thinking 1 currently has the strongest published LLMBoard score from this vendor.
The current public catalog contains 11 independently selectable Microsoft model products and 13 published versions.
Not necessarily. Vendor identifies the model developer; API providers are tracked separately.
1 current default versions are marked as open weight.