OpenAI compatible API · Attested · Public status
DigitalOcean Gradient AI performance
Review measured TTFT, effective throughput, uptime, and sampled model routes for DigitalOcean Gradient AI on TrustedRouter using metadata-only production probes.
Onebase URL to migrate
100sof models and routes
0prompt or output logs. Always.
DigitalOcean Gradient AIdigitalocean
36 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 | 1485 ms |
|---|---|
| p95 TTFT | 5665 ms |
| Effective throughput | 35 tok/s n=8 |
| Uptime | 100.00% |
Measured model routes
| Model | p50 TTFT | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|
| z-ai/glm-5 | 823 ms | — | 100.00% | — | 1 |
| moonshotai/kimi-k2.5 | 982 ms | — | 100.00% | — | 2 |
| minimax/minimax-m2.5 | 1021 ms | — | 100.00% | — | 1 |
| mistralai/ministral-3-14b-instruct | 1090 ms | — | 100.00% | — | 2 |
| moonshotai/kimi-k2.6 | 1103 ms | 39 tok/s n=2 | 100.00% | — | 1 |
| nvidia/nemotron-3-ultra-550b | 1108 ms | — | 100.00% | — | 2 |
| nvidia/nemotron-nano-12b-v2-vl | 1141 ms | — | 100.00% | — | 2 |
| z-ai/glm-5.1 | 1196 ms | — | 100.00% | — | 1 |
| deepseek/deepseek-v4-pro | 1302 ms | — | 100.00% | — | 1 |
| meta-llama/llama-4-maverick | 1424 ms | — | 100.00% | — | 4 |
| google/gemma-4-31b-it | 1485 ms | 72 tok/s n=2 | 100.00% | — | 1 |
| xiaomi/mimo-v2.5-pro | 2136 ms | 31 tok/s n=1 | 100.00% | — | 2 |
| qwen/qwen3.5-397b-a17b | 2161 ms | — | 100.00% | — | 3 |
| nvidia/nemotron-3-nano-omni | 2692 ms | — | 100.00% | — | 2 |
| deepseek/deepseek-v4-flash | 2965 ms | 7 tok/s n=2 | 100.00% | — | 5 |
| nvidia/nemotron-3-super-120b | 3019 ms | — | 100.00% | — | 3 |
| deepseek/deepseek-v3.2 | 3204 ms | — | 100.00% | — | 3 |
| z-ai/glm-5.2 | — | 12 tok/s n=1 | — | — | 0 |