Anthropic Just Dropped Claude Sonnet 5.5 (MAJOR UPGRADE)

Summary
Brock Mesarich reviews Anthropic’s announcement of Claude Sonnet 5.5, released just days after Claude Opus 5.5 as the second model in the Claude 5.5 family. He breaks down the official announcement blog post, covering pricing, performance benchmarks, industry feedback, and a coding speed comparison against Claude Sonnet 5. He also speculates on how this release positions Anthropic ahead of OpenAI’s upcoming DevDay.

What is shown

• [00:15] Anthropic’s official blog post (“Introducing Claude Sonnet 5.5”, dated September 28, 2026) alongside Mesarich’s digital whiteboard notes.
• [00:51] Announcement text noting that Claude Haiku 5.5 is slated to join the Claude 5.5 family in the coming weeks.
• [01:49] A side-by-side example comparing communication clarity between Claude Opus 5 and Claude Opus 5.5 on a code debugging explanation prompt.
• [02:25] Pricing table comparing Claude Sonnet 5.5 against Claude Opus 5.5 per million tokens.
• [03:23] Official benchmark performance table displaying scores across Sonnet 5.5, Sonnet 5, Opus 5.5, and GPT-6 Sol.
• [04:37] Testimonial quotes from Daniel Vogel (COO at Epic Games) and Sualeh Asif (Director of ML at SpaceXAI) evaluating Sonnet 5.5’s coding capabilities.
• [05:28] Side-by-side screen capture video comparing Claude Sonnet 5 and Claude Sonnet 5.5 executing the prompt: “A murmuration of 400 starlings in one HTML file”, showing Sonnet 5.5 generating code and running the canvas animation substantially faster.
• [06:03] An X post by @OpenAIDevs teasing OpenAI DevDay (“72 hours to OpenAI DevDay”).

Claims & numbers

• The presenter notes Claude Sonnet 5.5 was released shortly after Claude Opus 5.5.
• According to Anthropic’s announcement cited by the presenter, Sonnet 5.5 runs 30%+ faster and costs up to 30% less for most work compared to Claude Sonnet 5.
• Anthropic states Claude Haiku 5.5 will be released in the coming weeks.
• Pricing displayed:
  • Claude Sonnet 5.5: Cache reads $0.20 / 1M tokens, Cache writes $2.50 / 1M tokens, Input tokens $2 / 1M tokens, Output tokens $10 / 1M tokens.
  • Claude Opus 5.5: Cache reads $0.20 / 1M tokens, Cache writes $5 / 1M tokens, Input tokens $4 / 1M tokens, Output tokens $20 / 1M tokens.
• Benchmarks highlighted:
  • Agentic coding on Terminal Bench 4.0: Claude Sonnet 5.5 scores 70.6%, compared to 10.3% for Sonnet 5, 66.4% for Opus 5.5, and 49.3% for GPT-6 Sol.
  • CursorBench 4.0: Sonnet 5.5 scores 55.5% (High effort) vs 34.1% for Sonnet 5 and 57.0% for Opus 5.5.
  • FrontierCode 1.0 (Main): Sonnet 5.5 scores 46.2% at High effort (Sonnet 5: 42.4%, Opus 5.5: 54.4%, GPT-6 Sol: 49.3%) at roughly 1/15th the cost per task of Sonnet 5.
• The presenter predicts OpenAI will announce an AI personal assistant agent at DevDay, prompting a competitive response cycle from Anthropic.

Notable quotes

• [00:30] “Anthropic introduced Claude Sonnet 5.5, the second model in the Claude 5.5 family, as they just released Opus 5.5 just a few days ago...”
• [03:41] “5.5 is at 70.6%, whereas Sonnet 5 was 10.3%. So many people were complaining about the capabilities of Sonnet 5, so this does feel like a meaningful upgrade...”
• [04:43] “‘In Epic’s early testing, Claude Sonnet 5.5 cleared the same quality bar you’d expect from a higher-tier model...’” (quoting Daniel Vogel, COO at Epic Games).

Assessment
This is an independent creator review and commentary video analyzing Anthropic’s launch blog post and official demo footage. The presenter does not run original benchmark evaluations on-screen, relying entirely on Anthropic’s published data, quotes, and screen recording demonstrations.

Described by gemini-3.8-flash on 2026-09-29 from the video’s audio and frames.