GPT-6 Astra: providers compared
Every endpoint serving this model, side by side. ◆ marks the best value in each column — per model, never fleet-wide.
Current state
| State | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Azureazure | Healthy | 100%◆ (best) | 99.93% | 97.51% | 97.44% | 4.79s | 87.8s | 32 tps | 1.1M | $10.00 / $50.00 |
OpenAIopenai | Healthy | 100%◆ (best) | 99.99% | 99.66% | 99.64% | 4.02s | 71.5s | 34 tps | 1.1M | $10.00 / $50.00 |
OpenAIopenai/fast | Healthy | 100%◆ (best) | 100% | 99.99%◆ (best) | 99.99%◆ (best) | 4.95s | 54.7s | 37 tps | 1.1M | $20.00 / $100.00 |
OpenAIopenai/flex | Healthy | 100%◆ (best) | 99.95% | 99.96% | 99.97% | 3.36s◆ (best) | 42.6s | 43 tps◆ (best) | 1.1M | $5.00 / $25.00◆ (best) |
Azureazure/us | No data | — | 100% | 93.69% | 94.57% | 5.18s | 7.28s◆ (best) | 33 tps | 1.1M | $11.00 / $55.00 |
- Azure100%◆ (best)Healthy24h 99.93%
- 7d
- 97.51%
- 30d
- 97.44%
- p50
- 4.79s
- p99
- 87.8s
- Throughput
- 32 tps
- Context
- 1.1M
- $/Mtok
- $10.00 / $50.00
azure - OpenAI100%◆ (best)Healthy24h 99.99%
- 7d
- 99.66%
- 30d
- 99.64%
- p50
- 4.02s
- p99
- 71.5s
- Throughput
- 34 tps
- Context
- 1.1M
- $/Mtok
- $10.00 / $50.00
openai - OpenAI100%◆ (best)Healthy24h 100%
- 7d
- 99.99%◆ (best)
- 30d
- 99.99%◆ (best)
- p50
- 4.95s
- p99
- 54.7s
- Throughput
- 37 tps
- Context
- 1.1M
- $/Mtok
- $20.00 / $100.00
openai/fast - OpenAI100%◆ (best)Healthy24h 99.95%
- 7d
- 99.96%
- 30d
- 99.97%
- p50
- 3.36s◆ (best)
- p99
- 42.6s
- Throughput
- 43 tps◆ (best)
- Context
- 1.1M
- $/Mtok
- $5.00 / $25.00◆ (best)
openai/flex - No data24h 100%
- 7d
- 93.69%
- 30d
- 94.57%
- p50
- 5.18s
- p99
- 7.28s◆ (best)
- Throughput
- 33 tps
- Context
- 1.1M
- $/Mtok
- $11.00 / $55.00
azure/us
Measured 1m ago. Latency and throughput are OpenRouter's rolling 30-minute windows. ◆ marks the best value in each column among these endpoints.
Price vs latency
One dot per endpoint at the latest round · further left = faster, lower = cheaper · hover a dot for its tag
History
Availability
Axis starts at 90% so small differences stay visible; it expands to 0 when an endpoint falls below.
Time to first token
Logarithmic axis — the fleet spans two orders of magnitude. P50 is OpenRouter's rolling 30-minute window, averaged over the hour.
Throughput
Median tokens per second.