MiniMax M3: 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) | 96.88% | 99.01% | 90.52% | 1.37s | 56.4s | 66 tps | 524K | $0.300 / $1.20 | |
| Healthy | 100%◆ (best) | 99.88% | 99.75%◆ (best) | 98.90% | 1.43s | 9.60s | 26 tps | 524K | $0.280 / $1.10 | |
| Healthy | 100%◆ (best) | 99.39% | 99.39% | 99.51% | 3.16s | 11.1s | 46 tps | 1.0M | $0.240 / $0.960 | |
| Healthy | 100%◆ (best) | 97.89% | 98.34% | 99.00% | 1.76s | 2.98s | 126 tps | 1.0M | $0.750 / $3.00 | |
| Healthy | 100%◆ (best) | 99.64% | 99.51% | 99.63%◆ (best) | 2.08s | 14.9s | 65 tps | 1.0M | $0.300 / $1.20 | |
| Healthy | 100%◆ (best) | 99.81% | 99.58% | 98.44% | 793ms | 2.36s◆ (best) | 116 tps | 1.0M | $0.300 / $1.20 | |
| Healthy | 100%◆ (best) | 94.15% | 96.50% | 95.09% | 949ms | 5.46s | 87 tps | 524K | $0.300 / $1.20 | |
| Healthy | 99.89% | 96.97% | 97.41% | 98.81% | 723ms◆ (best) | 10.4s | 67 tps | 262K | $0.230 / $0.960◆ (best) | |
| Healthy | 98.80% | 99.74% | 99.62% | 99.34% | 1.17s | 10.5s | 76 tps | 524K | $0.300 / $1.20 | |
| Healthy | 98.77% | 99.06% | 98.55% | 99.06% | 1.46s | 8.84s | 61 tps | 1.0M | $0.300 / $1.20 | |
Togethertogether | Healthy | 97.53% | 99.47% | 99.52% | 99.54% | 1.22s | 9.00s | 28 tps | 524K | $0.300 / $1.20 |
Maramara | No data | — | 95.31% | 95.04% | 92.28% | 1.41s | 8.27s | 134 tps◆ (best) | 1.0M | $0.600 / $2.40 |
| No data | — | — | — | — | — | — | — | 256K | $0.255 / $1.02 | |
SambaNovasambanova | Degraded | — | 94.54% | 96.42% | 97.18% | 1.59s | 10.0s | 61 tps | 1.0M | $0.600 / $2.40 |
- AtlasCloud100%◆ (best)Healthy24h 96.88%
- 7d
- 99.01%
- 30d
- 90.52%
- p50
- 1.37s
- p99
- 56.4s
- Throughput
- 66 tps
- Context
- 524K
- $/Mtok
- $0.300 / $1.20
atlas-cloud/fp8fp8 - DeepInfra100%◆ (best)Healthy24h 99.88%
- 7d
- 99.75%◆ (best)
- 30d
- 98.90%
- p50
- 1.43s
- p99
- 9.60s
- Throughput
- 26 tps
- Context
- 524K
- $/Mtok
- $0.280 / $1.10
deepinfra/fp8fp8 - GMICloud100%◆ (best)Healthy24h 99.39%
- 7d
- 99.39%
- 30d
- 99.51%
- p50
- 3.16s
- p99
- 11.1s
- Throughput
- 46 tps
- Context
- 1.0M
- $/Mtok
- $0.240 / $0.960
gmicloud/fp8fp8 - ModelRun100%◆ (best)Healthy24h 97.89%
- 7d
- 98.34%
- 30d
- 99.00%
- p50
- 1.76s
- p99
- 2.98s
- Throughput
- 126 tps
- Context
- 1.0M
- $/Mtok
- $0.750 / $3.00
modelrun/fp4fp4 - Novita100%◆ (best)Healthy24h 99.64%
- 7d
- 99.51%
- 30d
- 99.63%◆ (best)
- p50
- 2.08s
- p99
- 14.9s
- Throughput
- 65 tps
- Context
- 1.0M
- $/Mtok
- $0.300 / $1.20
novita/fp8fp8 - Parasail100%◆ (best)Healthy24h 99.81%
- 7d
- 99.58%
- 30d
- 98.44%
- p50
- 793ms
- p99
- 2.36s◆ (best)
- Throughput
- 116 tps
- Context
- 1.0M
- $/Mtok
- $0.300 / $1.20
parasail/fp8fp8 - Venice100%◆ (best)Healthy24h 94.15%
- 7d
- 96.50%
- 30d
- 95.09%
- p50
- 949ms
- p99
- 5.46s
- Throughput
- 87 tps
- Context
- 524K
- $/Mtok
- $0.300 / $1.20
venice/fp8fp8 - CoreWeave99.89%Healthy24h 96.97%
- 7d
- 97.41%
- 30d
- 98.81%
- p50
- 723ms◆ (best)
- p99
- 10.4s
- Throughput
- 67 tps
- Context
- 262K
- $/Mtok
- $0.230 / $0.960◆ (best)
coreweave/fp4fp4 - Minimax98.80%Healthy24h 99.74%
- 7d
- 99.62%
- 30d
- 99.34%
- p50
- 1.17s
- p99
- 10.5s
- Throughput
- 76 tps
- Context
- 524K
- $/Mtok
- $0.300 / $1.20
minimax/fp8fp8 - StreamLake98.77%Healthy24h 99.06%
- 7d
- 98.55%
- 30d
- 99.06%
- p50
- 1.46s
- p99
- 8.84s
- Throughput
- 61 tps
- Context
- 1.0M
- $/Mtok
- $0.300 / $1.20
streamlake/fp8fp8 - Together97.53%Healthy24h 99.47%
- 7d
- 99.52%
- 30d
- 99.54%
- p50
- 1.22s
- p99
- 9.00s
- Throughput
- 28 tps
- Context
- 524K
- $/Mtok
- $0.300 / $1.20
together - Mara—No data24h 95.31%
- 7d
- 95.04%
- 30d
- 92.28%
- p50
- 1.41s
- p99
- 8.27s
- Throughput
- 134 tps◆ (best)
- Context
- 1.0M
- $/Mtok
- $0.600 / $2.40
mara - No data24h —
- 7d
- —
- 30d
- —
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 256K
- $/Mtok
- $0.255 / $1.02
morph/fp4fp4 - Degraded24h 94.54%
- 7d
- 96.42%
- 30d
- 97.18%
- p50
- 1.59s
- p99
- 10.0s
- Throughput
- 61 tps
- Context
- 1.0M
- $/Mtok
- $0.600 / $2.40
sambanova
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.
Throughput
Median tokens per second.