Term
Opus 4.6
Claude Opus 4.6 is Anthropic's flagship model from February 2026 — top tier of the Claude 4 family with a 1M-token context window, stronger coding and new effort controls.
Opus 4.6 — explained in more detail
Claude Opus 4.6 launched on February 5, 2026 and replaced Opus 4.5 as the top-tier model in the Claude 4 family. Compared to its predecessor it plans more carefully, sustains long agentic runs more reliably and operates more confidently in large codebases — with notable gains in code review and debugging. It posted the then-leading score on Terminal-Bench 2.0 for agentic coding and topped Humanity’s Last Exam.
Opus, Sonnet, Haiku — when to pick which?
Model selection in Claude Code (May 2026)
Anthropic tiers the Claude 4 family by use case. Opus 4.7 is the default pick for complex work with a 1M-token context window. Sonnet 4.6 remains the everyday workhorse — faster, cheaper, sufficient for most tasks. Haiku 4.5 delivers the fastest responses and is ideal for quick lookups or high-frequency calls.
Opus 4.6 is the immediate predecessor of the 4.7 default shown here — the selection logic (tier by quality vs. speed vs. cost) applies analogously.
What is new compared to Opus 4.5
- 1M-token context window (beta) — a fivefold jump from the 200k of Opus 4.5.
- Adaptive thinking: the model judges for itself how much extended thinking a task warrants; new effort controls give developers a dial between quality, speed and cost.
- Agent teams in Claude Code: multiple agents can be assembled into a team that collaborates on a single task.
- Workplace tools: substantial upgrades to Claude in Excel, plus Claude in PowerPoint as a research preview.
Where it fits
Opus stays the right pick when quality matters more than speed — hard coding problems, long agentic runs, deep research. For everyday interactive work Sonnet 4.6 remains the default. Pricing for Opus 4.6 is unchanged at $5 per million input tokens and $25 per million output tokens.
Discover more
AI Recap July 2026: Two Anthropic Models, a Government Review and the Open-Weight Pressure
July 2026 in AI: Sonnet 5 and Opus 5 from Anthropic, GPT-5.6 after a US review, Kimi K3 as open weights, and new GEO controls in Search Console.
GlossaryClaude Fable 5.1
Anthropic's flagship Claude Fable model, released on 1 September 2026. A frontier model for demanding reasoning, coding, research and long-running agents; 1M-token context, multimodal input, proprietary access via API and the Claude apps.
EncyclopediaComparing AI models — who builds what and how to choose
The major model families in 2026 at a glance. Who builds Claude, GPT, Gemini, Llama, Mistral, DeepSeek, Qwen — and which model to pick when.