Ollama cloud plan: Pro $20/mo (50x Free usage, 3 concurrent models). NOT toke...¶
Ollama cloud plan: Pro $20/mo (50x Free usage, 3 concurrent models). NOT token-metered — GPU time based. Larger models = more GPU time. Adam hit 90% with Hermes alone on glm-5.1:cloud. Strategy: offload simple crons to free models (Gemini Flash-Lite via AI Studio), keep glm-5.1 for complex tasks only. Free model tier: Gemini 2.5 Flash-Lite (AI Studio), NVIDIA NIM (40 RPM free), OpenRouter :free models, Cloudflare Workers AI ($0.017/M). Downgrade path: Email Triage, News, Action Log, Security, Branch Health, Bio Sync → free models. Keeps Workout/Podcast/Briefing on glm-5.1. Cost strategy at ~/projects/dual-agent-setup/COST-STRATEGY.md
Ollama cloud plan: Pro $20/mo (50x Free usage, 3 concurrent models). NOT token-metered — GPU time based. Larger models = more GPU time. Adam hit 90% with Hermes alone on glm-5.1:cloud. Strategy: offload simple crons to free models (Gemini Flash-Lite via AI Studio), keep glm-5.1 for complex tasks only. Free model tier: Gemini 2.5 Flash-Lite (AI Studio), NVIDIA NIM (40 RPM free), OpenRouter :free models, Cloudflare Workers AI ($0.017/M). Downgrade path: Email Triage, News, Action Log, Security, Branch Health, Bio Sync → free models. Keeps Workout/Podcast/Briefing on glm-5.1. Cost strategy at ~/projects/dual-agent-setup/COST-STRATEGY.md
Ollama cloud plan: Pro $20/mo (50x Free usage, 3 concurrent models). NOT token-metered — GPU time based. Larger models = more GPU time. Adam hit 90% with Hermes alone on glm-5.1:cloud. Strategy: offload simple crons to free models (Gemini Flash-Lite via AI Studio), keep glm-5.1 for complex tasks only. Free model tier: Gemini 2.5 Flash-Lite (AI Studio), NVIDIA NIM (40 RPM free), OpenRouter :free models, Cloudflare Workers AI ($0.017/M). Downgrade path: Email Triage, News, Action Log, Security, Branch Health, Bio Sync → free models. Keeps Workout/Podcast/Briefing on glm-5.1. Cost strategy at ~/projects/dual-agent-setup/COST-STRATEGY.md