Skip to content
Open Source
Models/replicate/meta-llama-3-70b-instruct

Llama 3 70B Instruct (Replicate)

Active
replicate/meta-llama-3-70b-instruct
Input / Prompt$0.65 / M
Output / Completion$2.75 / M
Context Window8K
Throughput Speed80 tps
3 Providers Serving This ModelDeterministic Failover Active
ProviderInput /MOutput /MLatency (TTFT)Speed (TPS)UptimeQuantizationZDRAction
REPLICATEPrimary Upstream
$0.65 / M$2.75 / M0.12s80 tps99.98%FP16NoConnect / Configure Provider →
OpenRouter (Edge Router)Multi-Region Backup
$0.65 / M$2.75 / M0.24s110 tps99.92%FP8YesConnect / Configure Provider →
Cerebras Cloud LPUInference Accelerator
$0.00 / M$0.00 / M0.04s2,100 tps99.85%INT8NoConnect / Configure Provider →