Meta › Llama 3.3 70B Instruct › Pricing
Llama 3.3 70B Instruct pricing
Cheapest host: DeepInfra · checked Oct 23, 2026
$0.10 / $0.32per 1M tokens in / out
#127on OpenRouter by tokens · 36.3B / week
Price sheetper 1M tokens
| Input | $0.10 | DeepInfra |
| Output | $0.32 | |
| Context · max output | 128K · 4K | tokens |
Where to buy9 providers · per 1M tokens
| Provider | Input | Output | Context |
|---|---|---|---|
| DeepInfra | $0.10 | $0.32 | 131K |
| Novita | $0.14 | $0.40 | 12K |
| AkashML | $0.20 | $0.52 | 131K |
| Parasail | $0.22 | $0.50 | 131K |
| Cloudflare | $0.29 | $2.25 | 24K |
| Groq | $0.59 | $0.79 | 131K |
| CoreWeave | $0.71 | $0.71 | 128K |
| $0.72 | $0.72 | 128K | |
| Together | $1.04 | $1.04 | 131K |
In Meta's line-upper 1M in / out
| Llama 3.2 3B | $0.15 / $0.15 | −3% |
| Llama 4 Scout | $0.10 / $0.30 | −3% |
| Llama 4 Scout 17B 16E | $0.10 / $0.30 | −3% |
| Llama 3.3 70B Instruct | $0.10 / $0.32 | this model |
| Llama 3.2 11B | $0.16 / $0.16 | +3% |
| Llama Guard 4 12B | $0.18 / $0.18 | +16% |
| Llama 3.1 8B | $0.22 / $0.22 | +42% |
Cheaper for the same workvs Llama 3.3 70B Instruct
| Gemma 3 27B unstatedOpen | -35% | score wins 59% | $0.08 / $0.16 |
| Qwen 3.5 Flash unstated | -27% | score wins 58% | $0.065 / $0.26 |
| Gemma 3 12B unstatedOpen | -60% | score wins 57% | $0.05 / $0.10 |
| GLM 4.7 Flash unstatedOpen | -6% | score wins 53% | $0.061 / $0.40 |
| GPT OSS 120B unstatedOpen | -61% | same score | $0.03 / $0.15 |
Pricing questions
How much does Llama 3.3 70B Instruct cost?›
$0.10 per 1M input tokens and $0.32 per 1M output tokens.
Is Llama 3.3 70B Instruct free to use?›
There is no free API tier, but the weights are open: you can run it on your own hardware for the cost of compute.
What is a cheaper alternative to Llama 3.3 70B Instruct?›
Gemma 3 27B at unstated effort costs -35% per task with a score wins 59% on Arena Text · overall.