MiMo-V2.6-Flash: 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) | 99.64% | 99.39% | 99.39% | 4.75s | 16.5s | 23 tps | 1.0M | $0.100 / $0.280 | |
| Healthy | 100%◆ (best) | 99.69% | 99.51%◆ (best) | 99.51%◆ (best) | 1.01s | 7.56s | 67 tps | 1.0M | $0.115 / $0.280 | |
| Healthy | 100%◆ (best) | 99.56% | 98.49% | 98.49% | 1.35s | 6.21s | 82 tps◆ (best) | 1.0M | $0.118 / $0.252 | |
| Healthy | 100%◆ (best) | 98.17% | 94.58% | 94.58% | 1.27s | 5.97s◆ (best) | 25 tps | 1.0M | $0.175 / $0.350 | |
| Healthy | 99.88% | 98.59% | 96.45% | 94.29% | 980ms◆ (best) | 7.99s | 52 tps | 1.0M | $0.140 / $0.280 | |
| Healthy | 99.76% | 99.19% | 98.88% | 99.09% | 2.96s | 77.6s | 44 tps | 1.0M | $0.140 / $0.280 | |
| Degraded | 98.17% | 96.14% | 96.20% | 96.20% | 2.66s | 96.5s | 28 tps | 1.0M | $0.140 / $0.280 | |
| Degraded | 96.62% | 86.54% | 87.10% | 87.10% | 3.33s | 7.13s | 43 tps | 1.1M | $0.140 / $0.280 | |
| No data | — | — | 98.66% | 98.66% | — | — | — | 1.0M | $0.120 / $0.280 | |
| No data | — | — | 97.88% | 97.88% | — | — | — | 1.0M | $0.040 / $1.28◆ (best) |
- Darkbloom100%◆ (best)Healthy24h 99.64%
- 7d
- 99.39%
- 30d
- 99.39%
- p50
- 4.75s
- p99
- 16.5s
- Throughput
- 23 tps
- Context
- 1.0M
- $/Mtok
- $0.100 / $0.280
darkbloom/fp4fp4 - InferenceNet100%◆ (best)Healthy24h 99.69%
- 7d
- 99.51%◆ (best)
- 30d
- 99.51%◆ (best)
- p50
- 1.01s
- p99
- 7.56s
- Throughput
- 67 tps
- Context
- 1.0M
- $/Mtok
- $0.115 / $0.280
inference-net/fp8fp8 - Io Net100%◆ (best)Healthy24h 99.56%
- 7d
- 98.49%
- 30d
- 98.49%
- p50
- 1.35s
- p99
- 6.21s
- Throughput
- 82 tps◆ (best)
- Context
- 1.0M
- $/Mtok
- $0.118 / $0.252
io-net/fp8fp8 - Venice100%◆ (best)Healthy24h 98.17%
- 7d
- 94.58%
- 30d
- 94.58%
- p50
- 1.27s
- p99
- 5.97s◆ (best)
- Throughput
- 25 tps
- Context
- 1.0M
- $/Mtok
- $0.175 / $0.350
venice/fp8fp8 - DeepInfra99.88%Healthy24h 98.59%
- 7d
- 96.45%
- 30d
- 94.29%
- p50
- 980ms◆ (best)
- p99
- 7.99s
- Throughput
- 52 tps
- Context
- 1.0M
- $/Mtok
- $0.140 / $0.280
deepinfra/fp8fp8 - Xiaomi99.76%Healthy24h 99.19%
- 7d
- 98.88%
- 30d
- 99.09%
- p50
- 2.96s
- p99
- 77.6s
- Throughput
- 44 tps
- Context
- 1.0M
- $/Mtok
- $0.140 / $0.280
xiaomi/fp8fp8 - Novita98.17%Degraded24h 96.14%
- 7d
- 96.20%
- 30d
- 96.20%
- p50
- 2.66s
- p99
- 96.5s
- Throughput
- 28 tps
- Context
- 1.0M
- $/Mtok
- $0.140 / $0.280
novita/fp8fp8 - GMICloud96.62%Degraded24h 86.54%
- 7d
- 87.10%
- 30d
- 87.10%
- p50
- 3.33s
- p99
- 7.13s
- Throughput
- 43 tps
- Context
- 1.1M
- $/Mtok
- $0.140 / $0.280
gmicloud/bf16bf16 - No data24h —
- 7d
- 98.66%
- 30d
- 98.66%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $0.120 / $0.280
inference-net/mxfp4mxfp4 - No data24h —
- 7d
- 97.88%
- 30d
- 97.88%
- p50
- —
- p99
- —
- Throughput
- —
- Context
- 1.0M
- $/Mtok
- $0.040 / $1.28◆ (best)
relace/mxfp4mxfp4
Measured 4m 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.