GLM 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Healthy | 100%◆ (best) | 99.99% | 99.99%◆ (best) | 99.97%◆ (best) | 1.98s | 12.3s | 50 tps | 203K | $1.00 / $3.20 | |
| Healthy | 99.41% | 99.90% | 99.80% | 99.74% | 2.74s | 11.6s | 43 tps | 198K | $0.600 / $1.92◆ (best) | |
| Healthy | 99.25% | 99.96% | 99.95% | 99.93% | 6.40s | 13.7s | 58 tps | 203K | $1.00 / $3.20 | |
| Healthy | 97.58% | 99.50% | 99.39% | 98.48% | 3.11s | 13.6s | 47 tps | 203K | $0.600 / $1.92◆ (best) | |
Amazon Bedrockamazon-bedrock | No data | — | 99.14% | 99.69% | 97.89% | 851ms◆ (best) | 4.00s◆ (best) | 110 tps◆ (best) | 203K | $1.00 / $3.20 |
| No data | — | — | — | — | — | — | — | 203K | $0.950 / $3.15 | |
| No data | — | 99.33% | 99.49% | 99.61% | 1.50s | 17.4s | 47 tps | 203K | $0.700 / $2.24 | |
| No data | — | — | — | 99.88% | — | — | — | 203K | $0.600 / $2.08◆ (best) | |
DigitalOceandigitalocean | No data | — | — | — | 97.63% | — | — | — | 64K | $1.00 / $3.20 |
| No data | — | 99.56% | 99.57% | 99.42% | 2.83s | 9.57s | 30 tps | 205K | $0.950 / $2.55 | |
| No data | — | 98.71% | 99.02% | 98.68% | 1.24s | 13.6s | 62 tps | 198K | $1.00 / $3.20 |
- Novita100%◆ (best)Healthy24h 99.99%
- 7d
- 99.99%◆ (best)
- 30d
- 99.97%◆ (best)
- p50
- 1.98s
- p99
- 12.3s
- Throughput
- 50 tps
- Context
- 203K
- $/Mtok
- $1.00 / $3.20
novita/fp8fp8 - StreamLake99.41%Healthy24h 99.90%
- 7d
- 99.80%
- 30d
- 99.74%
- p50
- 2.74s
- p99
- 11.6s
- Throughput
- 43 tps
- Context
- 198K
- $/Mtok
- $0.600 / $1.92◆ (best)
streamlake/fp8fp8 - Z.AI99.25%Healthy24h 99.96%
- 7d
- 99.95%
- 30d
- 99.93%
- p50
- 6.40s
- p99
- 13.7s
- Throughput
- 58 tps
- Context
- 203K
- $/Mtok
- $1.00 / $3.20
z-ai/fp8fp8 - GMICloud97.58%Healthy24h 99.50%
- 7d
- 99.39%
- 30d
- 98.48%
- p50
- 3.11s
- p99
- 13.6s
- Throughput
- 47 tps
- Context
- 203K
- $/Mtok
- $0.600 / $1.92◆ (best)
gmicloud/fp8fp8 - No data24h 99.14%
- 7d
- 99.69%
- 30d
- 97.89%
- p50
- 851ms◆ (best)
- p99
- 4.00s◆ (best)
- Throughput
- 110 tps◆ (best)
- Context
- 203K
- $/Mtok
- $1.00 / $3.20
amazon-bedrock - No data24h —
- 7d
- —
- 30d
- —
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 203K
- $/Mtok
- $0.950 / $3.15
atlas-cloud/fp8fp8 - No data24h 99.33%
- 7d
- 99.49%
- 30d
- 99.61%
- p50
- 1.50s
- p99
- 17.4s
- Throughput
- 47 tps
- Context
- 203K
- $/Mtok
- $0.700 / $2.24
baidu/fp8fp8 - No data24h —
- 7d
- —
- 30d
- 99.88%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 203K
- $/Mtok
- $0.600 / $2.08◆ (best)
deepinfra/fp4fp4 - No data24h —
- 7d
- —
- 30d
- 97.63%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 64K
- $/Mtok
- $1.00 / $3.20
digitalocean - No data24h 99.56%
- 7d
- 99.57%
- 30d
- 99.42%
- p50
- 2.83s
- p99
- 9.57s
- Throughput
- 30 tps
- Context
- 205K
- $/Mtok
- $0.950 / $2.55
siliconflow/fp8fp8 - No data24h 98.71%
- 7d
- 99.02%
- 30d
- 98.68%
- p50
- 1.24s
- p99
- 13.6s
- Throughput
- 62 tps
- Context
- 198K
- $/Mtok
- $1.00 / $3.20
venice/fp8fp8
Measured 5m 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.