70-80% of GitLab's LLM spend is currently with Anthropic.

However, they're re-routing their API requests regularly to the LLM provider that has the best coding or code summarization capabilities.

One year ago, Google was getting the most allocation with their Codey model. However, Gemini isn't great at writing code, so Anthropic is currently making the money.

Anthropic also provides a more stable LLM for enterprises. It seems that OpenAI, Gemini etcetera tend to drift as they update their LLMs. A…

Former semi engineer at Google and Groq says Nvidia's moat in training is unbreakable for the foreseeable future, but the market in fine-tuning and inference will be more competitive. AMD doesn't have a chance in training and for inference & fine-tuning, hyperscalers are moving to ASICs. (source: Tegus).