Llama 3.3 70 B Instruct pricing
Source: openrouter · updated 2026-11-18
Short answer: Llama 3.3 70 B Instruct costs $0.710 per million input tokens and $0.710 per million output tokens. That is a 3:1 blended rate of $0.710 per million tokens, ranking it #494 of 833 tracked models by input price.
Input / 1M
$0.710
Output / 1M
$0.710
Output ÷ input
1.00×
Context window
131,072
Price breakdown
| Metric | USD |
|---|---|
| Input, per 1M tokens | $0.710 |
| Output, per 1M tokens | $0.710 |
| Output ÷ input multiple | 1.00× |
| 3:1 blended, per 1M tokens | $0.710 |
| Context window | 131,072 tokens |
| Vendor | meta-llama |
What a typical workload costs
| Monthly volume | Input cost | Output cost | Total |
|---|---|---|---|
| 1M in / 250K out | $0.710 | $0.177 | $0.887 |
| 10M in / 2.5M out | $7.10 | $1.77 | $8.88 |
| 100M in / 25M out | $71.00 | $17.75 | $88.75 |
Other meta-llama models
| Model | Input | Output |
|---|---|---|
| Llama 3.1.8 B Instruct | $0.020 | $0.040 |
| Llama 3.2.1 B Instruct | $0.027 | $0.201 |
| Llama 3.2.3 B Instruct | $0.051 | $0.335 |
| Llama 4 Scout | $0.100 | $0.300 |
| Llama Guard 4 12 B | $0.180 | $0.180 |
Cheaper alternatives
Next cheaper option: GPT-4.1 Mini at $0.400 per 1M input tokens. Next step up: Qwen 3.5 122 B A10 B at $0.260 per 1M input tokens.
Spotted a price that is out of date? Submit a correction.