Anthropic launched Claude Sonnet 5.5 on September 28, 2026 — the second model in the Claude 5.5 family, six days after Opus 5.5. Sonnet 5.5 runs 30% faster than Sonnet 5 and costs up to 30% less for most work. In agentic build benchmarks, Sonnet 5.5 produced apps scoring level with Opus 5 while taking only 3.6 iterations per build (Opus 5 averaged 7.7). Fewest failed tool calls of any Anthropic model tested.
WHY IT MATTERS
The AI coding race is now about cost per completed task, not benchmark scores. Sonnet 5.5 makes routine agentic work dramatically cheaper. For most day-to-day Claude Code sessions — code review, refactors, writing tests, debugging — Sonnet 5.5 is now the correct default. Opus 5.5 stays for genuinely complex multi-hour work.
HOW TO USE IT
API model string: claude-sonnet-5-5. Drop-in replacement for claude-sonnet-5. Recommended for: standard coding tasks, tool use, moderate reasoning, cost-sensitive agentic workflows. For hardest tasks, use Opus 5.5 or Fable 5.1.
Anthropic launched Claude Opus 5.5 on September 22, 2026 — the first model in the new Claude 5.5 family. It matches Claude Fable 5.1 performance on most work at 40% lower cost than Opus 5. Pricing: $4 per million input tokens, $20 per million output tokens, $0.20 per million cache reads. 1M-token context window, 128K max output, always-on adaptive thinking. Available on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud, and Microsoft Foundry.
WHY IT MATTERS
Opus 5.5 collapses the price-performance frontier. Frontier-level capability (matching Fable 5.1) at Opus pricing tier makes agentic coding economical for continuous, multi-hour sessions. For builders: this is the model to default to for any complex agentic workflow starting Sept 22.
HOW TO USE IT
API model string: claude-opus-5-5. Migration from claude-opus-5 is drop-in — same schema, adaptive thinking now always on. Recommended for: multi-hour agentic sessions, large codebases, complex reasoning tasks. For simpler tasks, use Sonnet 5.5 (Sept 28).
Claude Code moved from public beta to general availability status. All features are now production-ready with enterprise SLA. Available on macOS, Linux, and Windows. Integrates with Claude Opus 5, Sonnet 5, Fable 5, and Haiku 4.5 model tiers.
WHY IT MATTERS
GA status signals Anthropic has committed to Claude Code as a durable product surface. Enterprise buyers now have SLA guarantees for procurement. Individual developers get stability commitments — no more breaking changes without deprecation cycles. This is the moment where Claude Code stops being 'the experimental CLI' and starts being 'the default agentic coding tool'.
HOW TO USE IT
Install: `npm install -g @anthropic/claude-code` (macOS/Linux) or via the Windows MSI installer. Existing beta users are auto-migrated. Check version: `claude --version` should show 1.0+ for GA release.
CLAUDE / BASH
# Install Claude Code GA releasenpm install -g @anthropic/claude-code# Verify GA version (should show 1.0.x)claude --version# Start an agentic session with Opus 5 (adaptive thinking default)claude --model opus-5 "Refactor this 40-file service..."
Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on September 1, 2026. Both are the same underlying model — Mythos 5.1 is the restricted-access version available through Project Glasswing, Fable 5.1 is the generally available version with additional safeguards. Same 1M-token context window, 128K max output, and same API pricing as their v5 predecessors. Positioned as Anthropic's most capable generally available model for demanding reasoning, long-running agents, coding, research, and document work.
WHY IT MATTERS
Fable 5.1 is the current top-tier commercial Anthropic model until Opus 5.5 launched three weeks later at similar performance for 40% less. If you were paying Fable pricing for tasks that don't strictly need Fable-level capability, Opus 5.5 (Sept 22) is now the cost-optimal choice. Fable 5.1 remains correct for the hardest reasoning tasks.
HOW TO USE IT
API model string: claude-fable-5-1. Mythos 5.1 is invite-only via Project Glasswing at anthropic.com/glasswing. For most users, Opus 5.5 (released Sept 22, 2026 at 40% lower cost) is now the better default.
Four focused micro-SaaS products targeting remote workers and nomads: an AI-powered nomad visa checker, a cost-of-living comparison tool with AI summaries, a coworking space finder with natural language search, and a freelance rate calculator with market benchmarking.
KEY LESSONS
— Build in public from day one — your first 200 users come from Twitter followers watching you build
— One tool, one problem, one price. No pricing tiers, no enterprise plans, no sales calls
— Claude API is so cheap that AI features are now table stakes, not premium features
Integrated Claude-powered AI writing assistance into ShipFast, his Next.js boilerplate for indie hackers. The AI feature generates landing page copy, email sequences, and changelog entries in the developer's brand voice. Added as an optional module, not a forced upsell.
KEY LESSONS
— Selling to developers means technical credibility first, then marketing
— One-time pricing outperforms subscriptions for bootstrapped tools — less churn anxiety, faster word-of-mouth
— The Claude API cost is so low it's basically a rounding error — don't overthink AI pricing