Meta: Llama 3.3 70B Instruct Performance
TrustedRouter performance signals and provider route posture for Meta: Llama 3.3 70B Instruct.
meta-llama/llama-3.3-70b-instruct
Measured performance
Continuously sampled p50/p95 time-to-first-token (TTFT), time-to-first-byte (TTFB), throughput, and success rate for Meta: Llama 3.3 70B Instruct — unsupported route and probe-configuration rows are separated from provider downtime, and no prompt or output content is stored.
| Provider | p50 TTFT | p95 TTFT | p50 TTFB | Throughput | Uptime | Config excluded | Samples |
|---|---|---|---|---|---|---|---|
| novita | 1106 ms | 9664 ms | 1105 ms | — | 66.67% | — | 3 |
| digitalocean | 1883 ms | 10564 ms | 1882 ms | — | 87.50% | — | 8 |
| crusoe | 3103 ms | 16009 ms | 3103 ms | — | 100.00% | — | 6 |
| makora | 3607 ms | 16435 ms | 3607 ms | — | 100.00% | — | 30 |
| together | 4483 ms | 19901 ms | 4483 ms | — | 100.00% | — | 16 |
| parasail | 5034 ms | 10344 ms | 5034 ms | — | 100.00% | — | 6 |
| tinfoil | 5333 ms | 13036 ms | 5333 ms | — | 100.00% | — | 21 |
Full provider & model leaderboard.
14 routes.
More routes give the auto router more room to fail over around provider 429 and 5xx responses.
Gateway overhead is measured separately.
Public status separates TLS/health overhead from full model latency so slow LLMs do not inflate the router metric.
Metadata rollups.
Status samples store latency, outcome, provider, model, route, cost, and region metadata only.
View public status or inspect provider routes.