Claude Sonnet 5.5: 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Anthropicanthropic | Healthy | 100%◆ (best) | 99.98% | 99.99% | 99.99% | 1.37s◆ (best) | 7.44s◆ (best) | 88 tps | 1.0M | $2.00 / $10.00◆ (best) |
Azureazure/global | Healthy | 100%◆ (best) | 99.78% | 99.86% | 99.86% | 1.95s | 17.8s | 106 tps | 1.0M | $2.00 / $10.00◆ (best) |
Claude Platform on AWSclaude-on-aws | Healthy | 100%◆ (best) | 99.96% | 99.93% | 99.93% | 2.70s | 10.3s | 79 tps | 1.0M | $2.00 / $10.00◆ (best) |
Googlegoogle-vertex/global | Healthy | 100%◆ (best) | 99.99% | 100%◆ (best) | 100%◆ (best) | 2.47s | 16.5s | 105 tps | 1.0M | $2.00 / $10.00◆ (best) |
Amazon Bedrockamazon-bedrock | No data | — | 99.95% | 100%◆ (best) | 100%◆ (best) | 2.66s | 12.9s | 99 tps | 1.0M | $2.00 / $10.00◆ (best) |
Azureazure/us | No data | — | — | — | — | — | — | — | 1.0M | $2.20 / $11.00 |
Googlegoogle-vertex/europe | No data | — | 99.99% | 100%◆ (best) | 100%◆ (best) | 2.87s | 9.51s | 98 tps | 1.0M | $2.20 / $11.00 |
Googlegoogle-vertex/us | No data | — | 99.93% | — | — | 4.04s | 16.2s | 121 tps◆ (best) | 1.0M | $2.20 / $11.00 |
- Anthropic100%◆ (best)Healthy24h 99.98%
- 7d
- 99.99%
- 30d
- 99.99%
- p50
- 1.37s◆ (best)
- p99
- 7.44s◆ (best)
- Throughput
- 88 tps
- Context
- 1.0M
- $/Mtok
- $2.00 / $10.00◆ (best)
anthropic - Azure100%◆ (best)Healthy24h 99.78%
- 7d
- 99.86%
- 30d
- 99.86%
- p50
- 1.95s
- p99
- 17.8s
- Throughput
- 106 tps
- Context
- 1.0M
- $/Mtok
- $2.00 / $10.00◆ (best)
azure/global - Claude Platform on AWS100%◆ (best)Healthy24h 99.96%
- 7d
- 99.93%
- 30d
- 99.93%
- p50
- 2.70s
- p99
- 10.3s
- Throughput
- 79 tps
- Context
- 1.0M
- $/Mtok
- $2.00 / $10.00◆ (best)
claude-on-aws - Google100%◆ (best)Healthy24h 99.99%
- 7d
- 100%◆ (best)
- 30d
- 100%◆ (best)
- p50
- 2.47s
- p99
- 16.5s
- Throughput
- 105 tps
- Context
- 1.0M
- $/Mtok
- $2.00 / $10.00◆ (best)
google-vertex/global - No data24h 99.95%
- 7d
- 100%◆ (best)
- 30d
- 100%◆ (best)
- p50
- 2.66s
- p99
- 12.9s
- Throughput
- 99 tps
- Context
- 1.0M
- $/Mtok
- $2.00 / $10.00◆ (best)
amazon-bedrock - No data24h —
- 7d
- —
- 30d
- —
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $2.20 / $11.00
azure/us - No data24h 99.99%
- 7d
- 100%◆ (best)
- 30d
- 100%◆ (best)
- p50
- 2.87s
- p99
- 9.51s
- Throughput
- 98 tps
- Context
- 1.0M
- $/Mtok
- $2.20 / $11.00
google-vertex/europe - No data24h 99.93%
- 7d
- —
- 30d
- —
- p50
- 4.04s
- p99
- 16.2s
- Throughput
- 121 tps◆ (best)
- Context
- 1.0M
- $/Mtok
- $2.20 / $11.00
google-vertex/us
Measured 3m 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.
Throughput
Median tokens per second.