Ollama
One-command local LLM runtime.
OPEN SOURCECPU-CAPABLE
Anthropic's fast/cheap tier for high-volume calls.
Claude Haiku 4.5 is the small, fast member of the Claude family at $1/$5 per 1M tokens with a 200K context window. It is built for high-volume, latency-sensitive workloads — classification, extraction, routing, simple chat — where the big models are overkill.
High-volume pipelines: tagging, extraction, summarisation, simple chat.
Local models genuinely compete at this tier — a 14–32B open model on Ollama covers many Haiku-class workloads for free.
One-command local LLM runtime.