Arcee AI: Coder-Large is a powerful 32-billion-parameter model, an offspring of Qwen 2.5-Instruct, specifically fine-tuned for coding tasks. It excels in multi-file refactoring and long diff reviews thanks to its generous 32k context window. The model understands over 30 programming languages, with a particular focus on TypeScript, Go, and Terraform, making it a versatile tool for developers. Internal benchmarks demonstrate significant performance gains (5–8 pt) over CodeLlama-34B-Python on HumanEval and competitive BugFix scores, partly due to a reinforcement pass that prioritizes compilable output. Coder-Large emits structured explanations alongside code blocks by default, making it ideal for educational tools and production copilot scenarios. With a maximum output of 4K tokens and competitive pricing ($0.50/$0.80 per 1M tokens input/output), it offers a cost-effective solution for scalable interactive coding. Access this STARTER tier model on Multi AI.
✅ Best For
🚀 Capabilities
❌ Limitations
Specifications
| Provider | arcee-ai |
| Context Window | 32,768 tokens |
| Max Output | 4,096 tokens |
| Minimum Plan | Balance |
Pricing
| Input Price | $0.5000 / 1M tokens |
| Output Price | $0.8000 / 1M tokens |
💡 With PRO subscription, cost is reduced by 20%