Head-to-head comparison
Phi 4 reasoning vision 15B (Microsoft) and Phi mini MoE instruct (Microsoft) compared on the ModelCap Index, API price, context window, provider availability, weight access and every public benchmark board they share. Figures are the same ones shown on the live rankings; nothing here is a hidden score.
Snapshot as of 22 September 2026
As of 22 September 2026, Phi mini MoE instruct holds the stronger ModelCap Index position (#126 vs #224).
Evidence: finetune of microsoft/Phi-4-reasoning · finetune (100%) · global-corpus-prior over 208 held-out anchors (68%); launch card against 4 resolved peers on 3 rows (32%); exceeds every named peer on 3 of 3 rows; optimism haircut 6.91 from cross-lab-probe; shrunk 27 up toward the measured corpus. Prices are the lowest listed API offer per million tokens observed on the OpenRouter catalogue.
Neither model has a published result on a public benchmark board ModelCap tracks yet.
Scores are the sources' own published figures for each model's best evaluated configuration; ModelCap never re-runs a benchmark.
Want a different pairing? Open the interactive comparison tool to swap either model for any current ranked language model.
Phi mini MoE instruct ranks higher on the ModelCap Index as of 22 September 2026: #126 against #224. That is a capability ranking built from public benchmark evidence with published uncertainty; whether it is "better" for you also depends on price, context and where you can run it. Their published uncertainty intervals overlap, so the rank difference alone does not establish a reliable capability advantage for your workload.
At least one of the two has no listed API price on ModelCap right now, so no price comparison is made.
Both publish a —-token context window.
The two models do not share a coding benchmark board on ModelCap yet, so no head-to-head coding score is published; the ModelCap Index position is the closest overall signal.
Phi 4 reasoning vision 15B: Open weights. Phi mini MoE instruct: Open weights. Open weights mean the checkpoint can be downloaded and self-hosted under its licence; API-only models are available solely through hosted endpoints.
ModelCap currently lists 0 API providers for Phi 4 reasoning vision 15B and 0 for Phi mini MoE instruct, from the OpenRouter catalogue snapshot the site serves; each model page lists the providers and their prices.
Every figure comes from the sealed ModelCap dataset published 22 September 2026; the page re-renders within a minute of each data refresh, and the ModelCap Index positions are the same ones shown on the live rankings.