zachz/llm-cost-benchmark

---
license: mit
task_categories:
- text-generation
language:
- en
tags:
- llm
- cost
- benchmark
- pricing
- api
pretty_name: LLM Cost Benchmark
size_categories:
- n<1K
---

# LLM Cost Benchmark

Token pricing and latency benchmarks across major LLM providers. Updated dataset for cost estimation, budget planning, and provider comparison.

## Fields

- `provider`: API provider (OpenAI, Anthropic, Google, Meta, Mistral)
- `model`: Model name
- `input_cost_per_1m`: Input cost per 1M tokens (USD)
- `output_cost_per_1m`: Output cost per 1M tokens (USD)
- `context_window`: Maximum context window size
- `max_output`: Maximum output tokens
- `median_latency_ms`: Median latency for 100-token response (ms)
- `p99_latency_ms`: P99 latency (ms)
- `date_benchmarked`: When the benchmark was run

## Usage

```python
from datasets import load_dataset
ds = load_dataset("zachz/llm-cost-benchmark")
```

## License

MIT