NEWMEN
← back to all models

Meta: Llama 3.1 8B Instruct

meta-llama/llama-3.1-8b-instruct

Metaopen-weight131K context6 providersIntelligence 54.5· est.

Cheapest provider

$0.02 / 1M

DeepInfra

Fastest provider (p95)

788 tok/s

Groq

Intelligence (estimated)

54.5

Family + generation + popularity + price tier

Per-provider performance

Latency / throughput / uptime / price measured across providers over the last 30 minutes of live traffic. This is what proves “sourced cheapest” — Atlas mode draws on these per call to serve the cheapest path that holds quality.

ProviderQuantInput $/1MOutput $/1MLatency p50 / p95Throughput p50 / p95Uptime 30mSuccess
DeepInfraq8· fp8$0.0200$0.0300169ms / 488ms35 / 59 tok/s99.97%98.1%
DeepInfrafull· bf16$0.0200$0.0500222ms / 697ms27 / 49 tok/s99.99%99.9%
Novitaq8· fp8$0.0200$0.0500621ms / 813ms42 / 147 tok/s99.88%85.4%
Groqundisclosed$0.0500$0.0800292ms / 733ms160.5 / 788.4 tok/s99.99%85.0%
Cloudflareq8· fp8$0.1520$0.2870649ms / 1510ms17 / 23 tok/s100.00%100.0%
WandBfull· bf16$0.2200$0.2200153ms / 224ms142 / 218 tok/s100.00%93.9%

“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).

Intelligence estimate

No public benchmark numbers indexed for this model yet, so the leaderboard score is derived from the catalogue data we sync: model family standing, generation, and live usage rank — applied identically to open- and closed-weight models. The heuristic ceiling sits below confirmed-benchmark frontier models so curated rankings stay clearly on top.

Want a curated score? File a benchmark report and we’ll add it to the next sync.

How Atlas mode sources Meta: Llama 3.1 8B Instruct

  • Strict mode — pin Meta: Llama 3.1 8B Instruct exactly and we pass it straight through, sourced from the cheapest provider above. The same model, no substitutions — currently DeepInfra at $0.02/1M.
  • Atlas mode — the default. Each call is auto-optimized for the cheapest path that holds quality, at least 5% off going direct from call one and climbing as it ramps. You always see which model served the call and exactly what you saved — thumbs-down anything you don’t like for a full refund.