OpenAI: gpt-oss-120b Performance
Compare measured TTFT, throughput, uptime, and route health for gpt-oss-120b across TrustedRouter providers using metadata-only production probes.
openai/gpt-oss-120b
Measured performance
Continuously sampled p50/p95 time-to-first-token (TTFT), effective throughput, and success rate for OpenAI: gpt-oss-120b. 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.
| Provider | p50 TTFT | p95 TTFT | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|
| databricks | 1811 ms | 3678 ms | — | 100.00% | — | 11 |
| fireworks | 730 ms | 730 ms | — | 100.00% | — | 1 |
| baseten | 870 ms | 1699 ms | — | 100.00% | — | 4 |
| cerebras | 881 ms | 1984 ms | — | 100.00% | — | 7 |
| cloudflare-workers-ai | 1067 ms | 1951 ms | — | 100.00% | — | 2 |
| io-net | 1316 ms | 1316 ms | — | 100.00% | — | 1 |
| novita | 1431 ms | 1431 ms | — | 100.00% | — | 1 |
| tinfoil | 1653 ms | 3842 ms | — | 100.00% | — | 6 |
| nscale | 1668 ms | 1668 ms | 60 tok/s n=2 | 100.00% | — | 1 |
| together | 1757 ms | 1757 ms | — | 100.00% | 1 unsupported_route |
1 |
| siliconflow | 1764 ms | 1764 ms | — | 100.00% | — | 1 |
| akashml | 1808 ms | 6805 ms | — | 100.00% | — | 5 |
| scaleway | 1966 ms | 2255 ms | — | 100.00% | — | 4 |
| parasail | 2519 ms | 2519 ms | — | 100.00% | — | 1 |
| nebius | 15279 ms | 15279 ms | — | 100.00% | — | 1 |
| near-ai | — | — | — | 0.00% | — | 2 |
Full provider & model leaderboard.
19 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.