Skip to content
SLM-Works

Qwen 2.5 (3B Instruct)

Qwen · ~3B parameters · Under 4B · Last reviewed 2026-03-20

Why we use it

Compact Qwen tier for multilingual SLM tracks when 7B is too heavy but Phi-class baselines need an alternative.

License summary

Apache 2.0 for several Qwen 2.5 open weights - some derivatives use different licenses; read the card per checkpoint.

Typical deployment profiles

  • Edge / low footprint
  • VPC, single GPU class

Focus tags

  • General
  • Multilingual
  • Code

Typical use cases

  • Multilingual edge bots
  • Student checkpoints
  • Low-footprint RAG

External references

Related SLM-Works services

← Back to foundation models catalog · Insights · Glossary

Contact to validate a model choice for your environment.