Llama 3.3 70 B pricing
Source: litellm · updated 2026-10-27
Short answer: Llama 3.3 70 B costs $0.200 per million input tokens and $0.500 per million output tokens. That is a 3:1 blended rate of $0.275 per million tokens, ranking it #236 of 704 tracked models by input price.
Price breakdown
| Metric | USD |
|---|---|
| Input, per 1M tokens | $0.200 |
| Output, per 1M tokens | $0.500 |
| Output รท input multiple | 2.50× |
| 3:1 blended, per 1M tokens | $0.275 |
| Context window | 131,072 tokens |
What a typical workload costs
| Monthly volume | Input cost | Output cost | Total | 1M in / 250K out | $0.200 | $0.125 | $0.325 |
|---|---|---|---|---|---|---|---|
| 10M in / 2.5M out | $2.00 | $1.25 | $3.25 | ||||
| 100M in / 25M out | $20.00 | $12.50 | $32.50 |
Other inferenceapis models
| Model | Input | Output |
|---|---|---|
| DeepSeek | $0.190 | $0.380 |
| DeepSeek Reasoner | $0.190 | $0.380 |
| Llama 3.1.8 B Instant | $0.200 | $0.500 |
| Llama 3.3 70 B Versatile | $0.200 | $0.500 |
| Llama 3.3 70 B Instruct Turbo | $0.200 | $0.500 |
Compare against
Cheaper alternatives
Next cheaper option: DeepSeek Flash Latest at $0.020 per 1M input tokens. Next step up: Qwen 3 VL 32 B Instruct at $0.160 per 1M input tokens.