Meta: Llama 3.3 70B Instruct
by Meta LlamaMeta: Llama 3.3 70B Instruct is a large language model from Meta Llama. It costs $0.220 per million input tokens and $0.500 per million output tokens. Its context window is 131K tokens.
- Input / 1M tokens
- $0.220
- Output / 1M tokens
- $0.500
- Cached input / 1M
- $0.110
- Context window
- 131K
On repeated prefixes
tokens
Who serves it cheapest
10 hosts serve Llama 3.3 70B Instruct. Same weights, same API - the price difference is pure margin and routing.
| Provider | Input / 1M | Output / 1M | Context | Throughput | Uptime 24h |
|---|---|---|---|---|---|
| DeepInfraCheapestfp8 | $0.100 | $0.320 | 131K | - | 99.7% |
| Novitabf16 | $0.135 | $0.400 | 12K | - | 94.8% |
| AkashMLfp8 | $0.200 | $0.520 | 131K | - | 99.9% |
| Parasailfp8 | $0.220 | $0.500 | 131K | - | 100.0% |
| SambaNova | $0.450 | $0.900 | 131K | - | 96.8% |
| Groq | $0.590 | $0.790 | 131K | - | 100.0% |
| CoreWeavefp16 | $0.710 | $0.710 | 128K | - | 99.3% |
| $0.720 | $0.720 | 128K | - | - | |
| Cloudflarefp8 | $0.293 | $2.25 | 24K | - | 99.8% |
| Together | $1.04 | $1.04 | 131K | - | 96.1% |
The spread between DeepInfra and Together is 6.7× for identical weights. Quantization and context limits differ, so check both columns before switching.
Benchmarks
Independent scores published alongside the catalogue.
- Intelligence index
- 7.7
- Coding index
- 11.9
About Llama 3.3 70B Instruct
The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model...
Specifications
| Model ID | meta-llama/llama-3.3-70b-instruct |
|---|---|
| Provider | Meta Llama |
| Context window | 131K tokens |
| Max output | 16K tokens |
| Input modalities | text |
| Output modalities | text |
| Knowledge cutoff | 2023-12-31 |
| Open weights | Yes - meta-llama/Llama-3.3-70B-Instruct |
| Released | December 6, 2024 |
Cheaper alternatives
Models that cost less than Llama 3.3 70B Instruct while keeping at least half its context window and every input modality it supports.
Frequently asked
How much does Meta: Llama 3.3 70B Instruct cost?
$0.220 per million input tokens and $0.500 per million output tokens. Cached input reads cost $0.110 per million tokens.
What is the context window of Meta: Llama 3.3 70B Instruct?
131K tokens, with up to 16K tokens of output per request.
Which provider serves Meta: Llama 3.3 70B Instruct cheapest?
DeepInfra at $0.100 per million input tokens - 6.7× cheaper than Together, the most expensive of the 10 hosts serving it.