Llama 4 Maverick: 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 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Googlegoogle-vertex/us-east5 | Healthy | 100%◆ (best) | 99.69% | 99.55% | 99.55% | 778ms | 2.35s | 105 tps◆ (best) | 524K | $0.350 / $1.15 |
| Healthy | 100%◆ (best) | 99.72% | 99.24% | 99.24% | 472ms | 2.03s◆ (best) | 75 tps | 1.0M | $0.270 / $0.850 | |
DigitalOceandigitalocean | Healthy | 98.72% | 99.73% | 99.71% | 99.71% | 865ms | 4.10s | 11 tps | 128K | $0.200 / $0.696◆ (best) |
| Degraded | 97.12% | 99.52% | 99.88%◆ (best) | 99.88%◆ (best) | 505ms | 3.24s | 68 tps | 524K | $0.350 / $1.00 | |
| Degraded | 92.88% | 99.65% | 99.26% | 99.26% | 402ms◆ (best) | 2.11s | 67 tps | 1.0M | $0.200 / $0.800◆ (best) |
- Google100%◆ (best)Healthy24h 99.69%
- 7d
- 99.55%
- 30d
- 99.55%
- p50
- 778ms
- p99
- 2.35s
- Throughput
- 105 tps◆ (best)
- Context
- 524K
- $/Mtok
- $0.350 / $1.15
google-vertex/us-east5 - Novita100%◆ (best)Healthy24h 99.72%
- 7d
- 99.24%
- 30d
- 99.24%
- p50
- 472ms
- p99
- 2.03s◆ (best)
- Throughput
- 75 tps
- Context
- 1.0M
- $/Mtok
- $0.270 / $0.850
novita/fp8fp8 - DigitalOcean98.72%Healthy24h 99.73%
- 7d
- 99.71%
- 30d
- 99.71%
- p50
- 865ms
- p99
- 4.10s
- Throughput
- 11 tps
- Context
- 128K
- $/Mtok
- $0.200 / $0.696◆ (best)
digitalocean - Parasail97.12%Degraded24h 99.52%
- 7d
- 99.88%◆ (best)
- 30d
- 99.88%◆ (best)
- p50
- 505ms
- p99
- 3.24s
- Throughput
- 68 tps
- Context
- 524K
- $/Mtok
- $0.350 / $1.00
parasail/fp8fp8 - DeepInfra92.88%Degraded24h 99.65%
- 7d
- 99.26%
- 30d
- 99.26%
- p50
- 402ms◆ (best)
- p99
- 2.11s
- Throughput
- 67 tps
- Context
- 1.0M
- $/Mtok
- $0.200 / $0.800◆ (best)
deepinfra/basefp8
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.