Llama 3.1 8B
Llama 3.1 8B Instant
Section titled “Llama 3.1 8B Instant”Provider: Groq (Meta) Context Window: 128,000 tokens Quality: Medium Price: $0.20 per 1M input tokens
Overview
Section titled “Overview”Llama 3.1 8B is the fastest model in the Llama family. Running on Groq’s LPU, it delivers near-instant responses for simple to moderate tasks.
Benchmarks
Section titled “Benchmarks”| Benchmark | Score | Percentile |
|---|---|---|
| MMLU | 68.4% | Top 35% |
| HumanEval | 62.0% | Top 35% |
| MATH | 45.1% | Top 40% |
| GPQA | 28.0% | Top 40% |
| HellaSwag | 85.6% | Top 30% |
| ARC-Challenge | 83.4% | Top 30% |
Source: Meta AI technical reports
Best For
Section titled “Best For”- Quick Q&A and chat
- Simple coding tasks
- Content classification
- High-volume, low-latency workloads
Pricing
Section titled “Pricing”| Tier | Input | Output |
|---|---|---|
| BYOK | $0.20/1M | $0.20/1M |
| Marketplace | $0.206/1M | $0.206/1M |
Routing Recommendation
Section titled “Routing Recommendation”- Accurate mode: Rarely selected
- Balanced mode: Good for simple prompts
- Cheap mode: Primary choice for Groq users
- Eco mode: Top choice - minimal compute, fastest response
curl http://127.0.0.1:8787/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{"model": "fastlane-auto", "messages": [{"role": "user", "content": "What is the capital of France?"}]}'