Qwen3 235B A22B Instruct 2507
qwen/qwen3-235b-a22b-2507
Compare with Qwen3 30B A3B Instruct 2507, Qwen3 Coder 480B A35B, Qwen3 Next 80B A3B Instruct
CompareQwen3 235B A22B Instruct 2507
| Provider | State | Uptime 5m | 24h | p50 latency | p99 | Throughput | Context | $/Mtok |
|---|---|---|---|---|---|---|---|---|
Alibabaalibaba | Healthy | 100% | 99.97% | 478ms | 2.04s | 41 tps | 131K | $0.150 / $0.598 |
AtlasCloudatlas-cloud/fp8fp8 | Healthy | 100% | 99.74% | 1.09s | 17.7s | 26 tps | 131K | $0.200 / $0.880 |
Googlegoogle-vertex/us-south1 | Healthy | 100% | 98.99% | 588ms | 2.50s | 24 tps | 262K | $0.250 / $1.00 |
Nebiusnebius/fp8fp8 | Healthy | 100% | 92.14% | 599ms | 10.9s | 36 tps | 262K | $0.200 / $0.600 |
StreamLakestreamlake | Healthy | 100% | 99.79% | 761ms | 11.8s | 32 tps | 128K | $0.210 / $0.840 |
Parasailparasail/fp8fp8 | Healthy | 99.90% | 98.83% | 556ms | 3.52s | 26 tps | 131K | $0.140 / $0.800 |
GMICloudgmicloud/fp8fp8 | Healthy | 99.80% | 99.61% | 1.32s | 3.94s | 26 tps | 262K | $0.350 / $1.40 |
DeepInfradeepinfra/fp8fp8 | Healthy | 99.33% | 96.26% | 404ms | 6.24s | 16 tps | 262K | $0.090 / $0.550 |
Venicevenice/fp8fp8 | Healthy | 99.13% | 94.70% | 625ms | 4.65s | 14 tps | 128K | $0.150 / $0.750 |
Novitanovita/fp8fp8 | Healthy | 98.17% | 97.99% | 697ms | 29.3s | 18 tps | 131K | $0.090 / $0.580 |
- Alibaba100%Healthy24h 99.97%
- p50
- 478ms
- p99
- 2.04s
- Throughput
- 41 tps
- Context
- 131K
- $/Mtok
- $0.150 / $0.598
alibaba - AtlasCloud100%Healthy24h 99.74%
- p50
- 1.09s
- p99
- 17.7s
- Throughput
- 26 tps
- Context
- 131K
- $/Mtok
- $0.200 / $0.880
atlas-cloud/fp8fp8 - Google100%Healthy24h 98.99%
- p50
- 588ms
- p99
- 2.50s
- Throughput
- 24 tps
- Context
- 262K
- $/Mtok
- $0.250 / $1.00
google-vertex/us-south1 - Nebius100%Healthy24h 92.14%
- p50
- 599ms
- p99
- 10.9s
- Throughput
- 36 tps
- Context
- 262K
- $/Mtok
- $0.200 / $0.600
nebius/fp8fp8 - StreamLake100%Healthy24h 99.79%
- p50
- 761ms
- p99
- 11.8s
- Throughput
- 32 tps
- Context
- 128K
- $/Mtok
- $0.210 / $0.840
streamlake - Parasail99.90%Healthy24h 98.83%
- p50
- 556ms
- p99
- 3.52s
- Throughput
- 26 tps
- Context
- 131K
- $/Mtok
- $0.140 / $0.800
parasail/fp8fp8 - GMICloud99.80%Healthy24h 99.61%
- p50
- 1.32s
- p99
- 3.94s
- Throughput
- 26 tps
- Context
- 262K
- $/Mtok
- $0.350 / $1.40
gmicloud/fp8fp8 - DeepInfra99.33%Healthy24h 96.26%
- p50
- 404ms
- p99
- 6.24s
- Throughput
- 16 tps
- Context
- 262K
- $/Mtok
- $0.090 / $0.550
deepinfra/fp8fp8 - Venice99.13%Healthy24h 94.70%
- p50
- 625ms
- p99
- 4.65s
- Throughput
- 14 tps
- Context
- 128K
- $/Mtok
- $0.150 / $0.750
venice/fp8fp8 - Novita98.17%Healthy24h 97.99%
- p50
- 697ms
- p99
- 29.3s
- Throughput
- 18 tps
- Context
- 131K
- $/Mtok
- $0.090 / $0.580
novita/fp8fp8
Measured 2m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.
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 and p99 are OpenRouter's rolling 30-minute windows, averaged over the hour.
Throughput
Median tokens per second.