Find a model

Browse the full catalog →

Free to start

MiniMax: MiniMax M3 Performance

Try MiniMax M3 on us. Starter credit covers your first requests, then one key reaches all 12 provider routes with automatic failover.

minimax/minimax-m3

open weights Performance

All models

AI IQ IQ 114 #67 public AI IQ rank for minimax-m3
View AI IQ profile

Measured performance

Continuously sampled p50/p95 time-to-first-token (TTFT), effective throughput, and success rate for MiniMax: MiniMax M3. Effective throughput uses provider-reported output tokens over complete request time. Unsupported route and probe-configuration rows are separated from provider downtime, and no prompt or output content is stored.

Providerp50 TTFTp95 TTFTEffective throughputUptimeConfig excludedAvailability samples
minimax 1207 ms 5637 ms 75 tok/s n=2 100.00% 1 fusion_panel_admission_failed 17
telnyx 600 ms 1285 ms 252 tok/s n=2 100.00% — 2
wandb 1295 ms 1295 ms 263 tok/s n=2 100.00% — 7
parasail 1329 ms 1329 ms 179 tok/s n=2 100.00% — 7
streamlake 1785 ms 2216 ms 107 tok/s n=2 100.00% — 2
siliconflow 1830 ms 1830 ms 138 tok/s n=2 100.00% — 1
boundless — — 87 tok/s n=1 — — 0
deepinfra — — 71 tok/s n=2 — — 0
featherless — — 202 tok/s n=2 — — 0
fireworks — — 356 tok/s n=2 — — 0
nebius — — 314 tok/s n=2 — — 0
novita — — 132 tok/s n=2 — — 0

Full provider & model leaderboard.

Provider diversity

12 routes.

More routes give the auto router more room to fail over around provider 429 and 5xx responses.

Streaming

Gateway overhead is measured separately.

Public status separates TLS/health overhead from full model latency so slow LLMs do not inflate the router metric.

Status

Metadata rollups.

Status samples store latency, outcome, provider, model, route, cost, and region metadata only.

Workspace access

Sign in

Choose a sign in method to access your TrustedRouter workspace.

New to TrustedRouter?

By continuing you agree to the terms of service and privacy policy.