Llama 3.3 70B Instruct
DeprecatedLlama 3.3 70B Instruct is Meta's language model with a 131K context window and up to 120K output tokens, available from 21 providers, starting at $0.12 / 1M input and $0.2 / 1M output. Meta's 70B instruction-tuned LLM from Llama 3.3, optimized for complex instruction-following and deployed across multiple cloud regions.
| Benchmarks | |
|---|---|
| Intelligence Index | #364 |
| Coding Index | #177 |
| Math Index | #242 |
| MMLU-Pro | #216 |
| GPQA | #392 |
| HLE | #491 |
| LiveCodeBench | #232 |
| AIME | #86 |
| IFBench | #203 |
| Time to First Token | #431 |
| SciCode | #354 |
| MATH-500 | #116 |
| AIME 2025 | #242 |
| LCR | #347 |
| TerminalBench Hard | #313 |
| TAU2 | #282 |
| Output TPS | #138 |
Capabilities
Input1/5
Text✓
Image·
Audio·
Video·
PDF·
Output1/5
Text✓
Image·
Audio·
Video·
Embedding·
Capabilities4/13
Reasoning·
Adaptive Reasoning·
Function Calling✓
Parallel Function Calling✓
Structured Outputs✓
Native JSON Schema✓
Web Search·
URL Context·
Computer Use·
Code Execution·
File Search·
Prompt Caching·
Assistant Prefill·
Pricing by Provider
US Dollar ($)
Per 1M tokens
| Provider | Standard | Batch | ||||
|---|---|---|---|---|---|---|
| Input $ / 1M | Output $ / 1M | Cache Read $ / 1M | Cache Write 5m $ / 1M | Input $ / 1M | Output $ / 1M | |
| $0.72 | $0.72 | N/A | N/A | $0.36 | $0.36 | |
| $0.71 | $0.71 | N/A | N/A | — | — | |
| $0.2 | $0.2 | N/A | N/A | — | — | |
| $0.5 | $1.50 | $0.5 | $0.5 | — | — | |
| $0.23 | $0.4 | N/A | N/A | — | — | |
| $0.9 | $0.9 | N/A | N/A | — | — | |
| $0.65 | $0.65 | N/A | N/A | — | — | |
| $0.59 | $0.79 | N/A | N/A | — | — | |
| $0.135 | $0.4 | N/A | N/A | — | — | |
| $0.74 | $0.74 | N/A | N/A | — | — | |
| $1.03 | $1.03 | N/A | N/A | — | — | |
| $0.12 | $0.3 | N/A | N/A | — | — | |
| $0.71 | $0.71 | N/A | N/A | — | — | |
| $0.12 | $0.3 | N/A | N/A | — | — | |
| $0.13 | $0.4 | N/A | N/A | — | — | |
| $0.135 | $0.4 | N/A | N/A | — | — | |
| $0.2 | $0.2 | N/A | N/A | — | — | |
| $0.71 | $0.71 | $0.71 | N/A | — | — | |
| $0.72 | $0.72 | N/A | N/A | — | — | |
| $0.67 | $0.67 | N/A | N/A | — | — | |
| $0.6 | $1.20 | N/A | N/A | — | — | |
| $0.9 | $0.9 | N/A | N/A | — | — | |
| $0.71 | $0.71 | N/A | N/A | — | — | |
Cost Calculator
US Dollar ($)
Preset:
Cheapest Instances to Run It
Cloud GPU instances that can host Llama 3.3 70B Instruct, ranked by cheapest on-demand price. The model needs about 168 GB of GPU memory at FP16 precision (estimated from its parameter count), so treat the fit as guidance rather than a guarantee.
All clouds
FP16 (full precision)
US Dollar ($)
Instance | Cloud | GPU | VRAM | Price | Cheapest region | |
|---|---|---|---|---|---|---|
| Standard_NP40s |  Azure | 4× AMD Alveo U250 FPGA (64GB) | 256 GB | $6.60/hr | ||
| g2-standard-96 |  GCP | 8× nvidia-l4 | 192 GB | $7.98/hr | ||
| g7e.12xlarge |  AWS | 2× RTX PRO Server 6000 | 192 GB | $8.29/hr | ||
Versions
| Version | Released | Context | Input / 1M | Output / 1M | Status |
|---|---|---|---|---|---|
| Llama 3.3 70B Instruct | 131K | $0.120 | $0.200 | Current | |
| Llama 3.2 3B Instruct | 131K | $0.015 | $0.020 | Deprecated | |
| Llama 3.2 1B Instruct | 131K | $0.020 | $0.020 | Deprecated | |
| Llama 3.2 11B | 128K | $0.160 | $0.160 | Available | |
| Llama 3.1 405B Instruct | 131K | $0.120 | $0.300 | Deprecated | |
| Llama 3.1 8B Instruct | 200K | $0.020 | $0.030 | Deprecated | |
| Llama 3.1 70B Instruct | 131K | $0.100 | $0.100 | Deprecated | |
| Llama 3.1 70B | 128K | $0.360 | $0.360 | Available | |
| Llama 3.1 8B | 131K | $0.030 | $0.050 | Available | |
| Llama 3 70B Instruct | 131K | $0.120 | $0.300 | Deprecated | |
| Llama 3 8B Instruct | 32K | $0.030 | $0.040 | Available |