# Llama 3.2 3B Instruct API prices

Llama 3.2 3B Instruct (llama) — 131k context, 118k max output.
9 metered per-token offers from 10 provider(s), in USD per 1M tokens.

| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context |
| --- | --- | --- | --- | --- |
| Inference | $0.02 | $0.02 | — | 16k |
| DevPass (LLM Gateway) | $0.03 | $0.05 | — | 33k |
| LLM Gateway | $0.03 | $0.05 | — | 33k |
| NovitaAI | $0.03 | $0.05 | — | 33k |
| NanoGPT | $0.0306 | $0.0493 | $0.0153 | 131k |
| Kilo Gateway | $0.05 | $0.33 | — | 131k |
| OpenRouter | $0.05 | $0.33 | — | 131k |
| Cloudflare Workers AI | $0.0509 | $0.335 | — | 80k |
| Pioneer | $0.1 | $0.335 | $0.1 | 131k |

## Summary

- Cheapest input: $0.02 per 1M tokens (Inference)
- Cheapest output: $0.02 per 1M tokens (Inference)
- First-party: not listed separately
- Free offers: Nvidia
- Subscription plans (not per-token): none
- Released: 2024-09-18
- Inputs: pdf, text

HTML page: https://llmprice.gitlab.io/models/llama-3-2-3b-instruct/
Full dataset: https://llmprice.gitlab.io/data/catalog.json