Claude Model Family Reference
Comprehensive comparison of Anthropic's Claude model tiers: Opus 4.6, Sonnet 4.6, and Haiku 4.5. Context windows, capabilities, pricing, and recommended use cases from BNL's production experience.
Claude Model Comparison
Anthropic's Claude model family powers the majority of Blue Note Logic's AI-assisted workflows, from code generation to document analysis. This reference covers the current production models as of 2026.
Model Comparison
| Model | Strengths | Context | Best For |
|---|---|---|---|
| Claude Opus 4.6 | Highest reasoning, complex analysis | 200K tokens | Architecture decisions, complex code review, nuanced analysis |
| Claude Sonnet 4.6 | Best balance of speed and capability | 200K tokens | Code generation, document processing, general-purpose tasks |
| Claude Haiku 4.5 | Fastest response, lowest cost | 200K tokens | Classification, extraction, high-volume processing, real-time chat |
Key Capabilities
- Extended Thinking — Opus and Sonnet support chain-of-thought reasoning for complex multi-step problems
- Tool Use — All tiers support function calling for API integration and agentic workflows
- Vision — Image understanding across all model tiers
- Code Generation — Production-quality code in 50+ programming languages
BNL Usage Patterns
- Development: Claude Code (Opus 4.6) for architecture and complex implementations; Sonnet 4.6 for routine coding tasks
- Document Analysis: Sonnet 4.6 for processing legal and technical documents at scale
- Real-time Features: Haiku 4.5 for chat interfaces and high-throughput classification
Related Resources
dobetter-norge-v2
Domain-specific LLM fine-tuned for Norwegian legal text. Built on Qwen 2.5 7B with QLoRA 4-bit quantization, trained on 2,847 legislative documents and 7,603 expert Q&A pairs.
View details →
Qwen 2.5 7B Base Model
The base model architecture behind dobetter-norge-v2. Alibaba's Qwen 2.5 7B offers strong multilingual performance and serves as an excellent foundation for domain-specific fine-tuning.
Access →