[BUG] NIVIDIA NIM provider not currently supporting mini max 3 or problem with the provided model id

[mikmaneggahommie]

OmniRoute Version

3.8.12

Installation Method

Electron desktop app

Operating System

macOS

OS Version

macos monetery

Node.js Version

v24.16.0

Provider(s) Involved

NIVIDIA NIM

Model(s) Involved

nvidia/minimaxai/minimax-m3

Client Tool

claude code

Description

so when testing if the minimax m3 model works which is hosted by the NIVIDIA NIM , it returns
✕
[nvidia/minimaxai/minimax-m3] [404]: 404 page not found (reset after 2m) ,

i think they aren't supporting it yet , i also tried adding a custom model id for it , it still dont work , but other models work such as nvidia/moonshotai/kimi-k2.6 and nvidia/minimaxai/minimax-m2.7

Steps to Reproduce

just search for minimax , then click on the m3 version

Expected Behavior

it should work like the rest

Actual Behavior

it doesnt work

Test Impact

Unsure

Error Logs / Output

✕
[nvidia/minimaxai/minimax-m3] [404]: 404 page not found (reset after 2m)

Screenshots

Additional Context

i think it isn't supported yet

Validation Plan

No response

[diegosouzapw]

Thanks @mikmaneggahommie — you were right, NVIDIA NIM doesn't host minimaxai/minimax-m3 yet (hence the 404, while minimax-m2.7 works). Fixed in release/v3.8.14 by removing minimaxai/minimax-m3 from the NVIDIA tier so it no longer shows up as a selectable-but-broken model — MiniMax M3 stays available on the tiers that actually serve it. We'll re-add it to NVIDIA once they do. Will be in the next release (v3.8.14).