Models

Anthropic Makes Claude Sonnet 4.6 the Default Free and Pro Model

On February 17, 2026, Anthropic launched Claude Sonnet 4.6—its most capable Sonnet yet—as the default model for Free and Pro plans on Claude.ai and Claude Cowork. The release upgrades coding, computer use, long-context reasoning, agent planning, and knowledge work with a 1M-token context window in beta, while keeping Sonnet 4.5 pricing at $3/$15 per million tokens.

Tech Insights Reporter Feb 17, 2026 5 min read San Francisco, CA

TLDR

Anthropic released Claude Sonnet 4.6 on February 17, 2026, calling it the most capable Sonnet yet and making it the default on Free and Pro plans in claude.ai and Claude Cowork. Pricing stays at $3/$15 per million tokens (same as Sonnet 4.5). Early Claude Code users preferred it over Sonnet 4.5 about 70% of the time—and even preferred it to Opus 4.5 (November 2025) 59% of the time.

What shipped

From Anthropic’s primary announcement:

Positioning: Full upgrade across coding, computer use, long-context reasoning, agent planning, knowledge work, and design.
Context: 1M token context window in beta; described as enough for entire codebases, long contracts, or many research papers in one request, with improved long-horizon reasoning (e.g., Vending-Bench Arena strategy of early capacity investment then profit pivot).
Default product role: New default for Free and Pro on Claude.ai and Claude Cowork; free tier upgraded with file creation, connectors, skills, and compaction.
API / platform: Available on all Claude plans, Claude Code, API (claude-sonnet-4-6), and major cloud platforms; adaptive thinking, extended thinking, and context compaction (beta); web search/fetch code-filter improvements; several tools marked generally available.
Safety: System card evaluations; researchers described broadly warm/prosocial character, strong safety behaviors, no major high-stakes misalignment flags; improved prompt-injection resistance vs Sonnet 4.5, similar to Opus 4.6.
Computer use: Continued OSWorld gains; early users reported near human-level performance on complex spreadsheets and multi-step web forms (still lagging top human operators).

Early preference and partner signal

Anthropic reported Claude Code early testers preferred Sonnet 4.6 over Sonnet 4.5 ~70% and over Opus 4.5 59%, citing less overengineering, better instruction following, fewer false success claims, and stronger multi-step follow-through. Launch partners (Databricks, Replit, Cursor, GitHub, Cognition, and others) highlighted near-Opus performance on document/enterprise workloads at Sonnet cost.

Why this story matters

Twelve days after Opus 4.6, Anthropic put near-frontier capability on the default Free/Pro path—not only paid Max/API power users. That compresses the gap between “best model” and “what most people actually run,” and pressures rivals on price-performance for coding agents and computer use.

Sources

Anthropic: “Introducing Claude Sonnet 4.6” (anthropic.com/news/claude-sonnet-4-6, February 17, 2026). Primary.
Claude Sonnet 4.6 System Card (February 17, 2026).
Snowflake Cortex same-day availability note (February 17, 2026).

Featured Image Alt Text

Claude Sonnet 4.6 product graphic with Free/Pro default badge and 1M context window indicator.