Which open models should I use?

There’s no single right answer! Here’s a curated list based on Fireworks internal testing, community feedback, and external benchmarks. We recommend using it as a starting point, and we will update it regularly as new models emerge.

  • Model sizes are marked as Small, Medium, or Large
  • For best latency, use small or medium models. For best quality, use large models or fine-tune medium or small models.
  • You can explore all models in the Fireworks Model Library
Use CaseRecommended Models
Code generation & reasoningDeepseek R1 05/28, Deepseek V3 03-24 (Large)
Qwen2.5-32B-Coder (Medium)
Code completion & bug fixingQwen2.5-32B-Coder (Medium)
Qwen3 14B, Qwen3 7B (Small)
General reasoning & planningDeepSeek R1 05/28, Deepseek V3 03-24 (Large)
Qwen2.5-72B-Instruct, Llama 3.3 70B (Medium)
Function calling & tool useQwen3 235B A22B (Large)
Qwen 3 Family Models (Large/Medium/Small)
Long context & summarizationLlama 4 Maverick (Large)
Llama 4 Scout (Medium)
Vision & document understandingLlama 4 Maverick (Large)
Qwen2.5-VL 32B Instruct, Qwen2.5-VL 72B Instruct, Llama 4 Scout (Medium)
Qwen2.5-VL 3-7B (Small)
Low-latency NLU & extractionLlama 3.1 8B, Llama 3.2 3B, Llama 3.2 1B, Qwen3 7B (Small)

Last updated: June 30, 2025