Large language models and the platforms that power them.
Meta's open-weight instruction-following model. Available in 8B and 70B parameter sizes.
A 7B parameter model with strong performance on math and coding benchmarks.
Mixture-of-experts model. Only 12B active parameters per token, with 46B total.
Alibaba's multilingual LLM family with strong code and math capabilities.
Anthropic's most intelligent model. Excellent for complex reasoning and long-form content.
Google's flagship model with a 1M token context window. Excels at long-document reasoning.
OpenAI's most capable language model. Multimodal, 128k context window.