Start Free
puq/code is here. Try it on puq.sh
Integrations
Templates Apps Models Blog Pricing

Llama 3.3 70B Instruct FP8 Fast

Llama 3.3 70B quantized to fp8 precision, optimized to be faster.

Context
24K tokens
Input
$0.3663/1M
Output
$2.82/1M

Model facts

Provider
Meta
Task
Text Generation
Context length
24,000 tokens
Request format
JSON
Last updated
31 Aug 2026
Model code puqai/meta/llama-3.3-70b-instruct-fp8-fast

Pricing

Rate Price Per
Input $0.3663 1M
Output $2.82 1M

Capabilities

  • Chat
  • Streaming
  • Tools
  • Cloudflare
  • Hosted
  • LLM
Route AI with confidence

Ready to use Llama 3.3 70B Instruct FP8 Fast?