OpenAI compatible API · Attested · Public status
Together performance
Review measured TTFT, TTFB, effective throughput, uptime, and sampled model routes for Together on TrustedRouter using metadata-only production probes.
Onebase URL to migrate
100sof models and routes
0prompt or output logs. Always.
Togethertogether
35 samples
Continuously sampled provider performance. TrustedRouter reports unsupported route and probe-configuration rows separately from provider downtime. Prompt and output content is not stored.
| p50 TTFT | 1742 ms |
|---|---|
| p95 TTFT | 9245 ms |
| p50 TTFB | 2519 ms |
| Effective throughput | 57 tok/s n=11 |
| Uptime | 97.14% |
Measured model routes
| Model | p50 TTFT | p50 TTFB | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|
| moonshotai/kimi-k3 | 768 ms | 768 ms | 42 tok/s n=1 | 100.00% | — | 2 |
| google/gemma-3n-e4b-it | 927 ms | 927 ms | — | 100.00% | — | 1 |
| nvidia/nemotron-3-ultra-550b-a55b | 1191 ms | 1190 ms | 20 tok/s n=1 | 100.00% | — | 1 |
| qwen/qwen-2.5-7b-instruct | 1391 ms | 1391 ms | — | 100.00% | — | 8 |
| z-ai/glm-5.2 | 1662 ms | 1662 ms | — | 100.00% | — | 4 |
| thinkingmachines/inkling | 1742 ms | 1742 ms | 76 tok/s n=2 | 100.00% | — | 3 |
| moonshotai/kimi-k2.7-code | 2362 ms | 2361 ms | — | 100.00% | — | 7 |
| openai/gpt-oss-20b | 2598 ms | 2597 ms | — | 100.00% | — | 1 |
| openai/gpt-oss-120b | 2690 ms | 2690 ms | 82 tok/s n=2 | 100.00% | — | 3 |
| moonshotai/kimi-k2.6 | 3288 ms | 3288 ms | — | 100.00% | — | 3 |
| meta-llama/llama-3.3-70b-instruct | 3627 ms | 3626 ms | — | 100.00% | — | 1 |
| deepseek/deepseek-v4-flash-0731 | — | — | 57 tok/s n=2 | — | 4 probe_config_error |
0 |
| deepseek/deepseek-v4-pro | — | — | 37 tok/s n=1 | — | 5 probe_config_error |
0 |
| google/gemma-4-31b-it | — | — | 57 tok/s n=1 | — | 3 probe_config_error |
0 |
| minimax/minimax-m3 | — | — | 38 tok/s n=1 | — | 2 probe_config_error |
0 |
| qwen/qwen3.5-9b | — | — | — | 0.00% | 1 probe_config_error |
1 |