Complete Termux guide for 2026
Install:
pkg install ollama
ollama pull qwen2.5:0.6b
ollama serve
API:
curl http://localhost:11434/api/generate -d '{"model":"qwen2.5:0.6b","prompt":"Hello"}'
Use cases:
- Offline chat
- Local code assistance
- Privacy-first research
- Zero API costs
Limitations:
- 0.6B: fast but limited reasoning
- 4B+: better but slow on CPU
- No GPU acceleration on most Androids
Top comments (0)