Cost to run
Qwen 1.5 MoE A2.7B
alibaba/qwen-1.5-moe-a2.7b
- Family
- Qwen 1.5
- Context
- 32,768 tokens
Qwen 1.5 MoE A2.7B can be run self-hosted (rent a GPU + run vLLM/TGI) or through a serverless API (pay per token). Live pricing comparisons:
Self-hosted
Rent a GPU from Lambda, RunPod, Vast.ai, CoreWeave. Full control.
Serverless
Pay per token via Together, Fireworks, DeepInfra, Groq.
Need to benchmark this model against another? Try the calculator or see where it ranks on the InferenceScore leaderboard.