Llama 3.1 8B Instruct
meta-llama/llama-3.1-8b-instructInputOutput
Compare its 5 providersIncident RSSCompare with Llama 3.1 70B Instruct, Llama 3.2 3B Instruct, Llama 3.2 1B Instruct
CompareLlama 3.1 8B Instruct
| State | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Healthy | 100% | 100% | 100% | 99.96% | 274ms | 2.21s | 125 tps | 131K | $0.220 / $0.220 | |
| Healthy | 99.97% | 99.94% | 99.62% | 99.27% | 837ms | 3.36s | 28 tps | 131K | $0.020 / $0.040 | |
| Healthy | 99.89% | 99.76% | 99.26% | 98.54% | 526ms | 2.88s | 97 tps | 16K | $0.020 / $0.050 | |
Groqgroq | Healthy | 99.65% | 99.80% | 99.72% | 99.66% | 122ms | 881ms | 370 tps | 131K | $0.050 / $0.080 |
| No data | — | 99.60% | 99.09% | 98.66% | 501ms | 1.58s | 19 tps | 32K | $0.152 / $0.287 |
- CoreWeave100%Healthy24h 100%
- 7d
- 100%
- 30d
- 99.96%
- p50
- 274ms
- p99
- 2.21s
- Throughput
- 125 tps
- Context
- 131K
- $/Mtok
- $0.220 / $0.220
coreweave/bf16bf16 - DeepInfra99.97%Healthy24h 99.94%
- 7d
- 99.62%
- 30d
- 99.27%
- p50
- 837ms
- p99
- 3.36s
- Throughput
- 28 tps
- Context
- 131K
- $/Mtok
- $0.020 / $0.040
deepinfra/fp8fp8 - Novita99.89%Healthy24h 99.76%
- 7d
- 99.26%
- 30d
- 98.54%
- p50
- 526ms
- p99
- 2.88s
- Throughput
- 97 tps
- Context
- 16K
- $/Mtok
- $0.020 / $0.050
novita/fp8fp8 - Groq99.65%Healthy24h 99.80%
- 7d
- 99.72%
- 30d
- 99.66%
- p50
- 122ms
- p99
- 881ms
- Throughput
- 370 tps
- Context
- 131K
- $/Mtok
- $0.050 / $0.080
groq - No data24h 99.60%
- 7d
- 99.09%
- 30d
- 98.66%
- p50
- 501ms
- p99
- 1.58s
- Throughput
- 19 tps
- Context
- 32K
- $/Mtok
- $0.152 / $0.287
cloudflare/fp8fp8
Measured 5m 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 is OpenRouter's rolling 30-minute window, averaged over the hour.
Throughput
Median tokens per second.
Daily availability
One row per endpoint · one dot per day · last 90 days.
worst hour ≥ 95%90–95%< 90%no data
Colour by each endpoint's worst hour of each UTC day · 5 endpoints · 47 days with data, 20 endpoint-days below 90%.