Author of "Who makes money from open-source AI models?"
How this journalist typically writes
Darko as author
GPT-5 Mini offers best concurrency safeguards at lowest cost ($0.05), while Claude Haiku 4.5 provides fastest execution with zero tool-calling failures, and GLM-4.6 delivers multi-file architecture but requires disabling reasoning for tool calls.
“Author of "Mini Models Battle: Claude Haiku 4.5 vs GLM-4.6 vs GPT-5 Mini"”
Inference providers, not model creators, capture the majority of revenue from open-source AI models through competitive hardware infrastructure and API pricing.
“Author of "Who makes money from open-source AI models?"”