Claude Opus 5.5 Can Do More Than You Think...

Summary
The presenter provides an overview of Anthropic's Claude Opus 5.5 release, reviewing its benchmark performance and cost reductions compared to previous models. He then demonstrates a hands-on workflow using Claude Desktop alongside the Higgsfield MCP connector to programmatically automate and edit motion graphics directly inside Adobe After Effects.

What is shown

• [00:00] Anthropic’s launch page for Claude Opus 5.5 (dated September 22, 2026) and community demo showcases (Three.js Spider-Man clone, motion graphics showreels, and game prototypes).
• [00:46] Official Anthropic benchmark comparison table across Opus 5.5, Fable 5.1, Opus 5, GPT-6 Astra, and GPT-5.6 Sol.
• [01:21] Artificial Analysis leaderboard showing Opus 5.5 scoring 58 on the Intelligence Index.
• [01:44] An existing mobile-format explainer animation playing inside Adobe After Effects.
• [02:41] The Higgsfield plugin and MCP bridge integration website for Adobe software (After Effects, Premiere Pro, Photoshop).
• [04:47] Presenter prompts Claude via the desktop app to modify the After Effects project's brand colors (cyan and purple) and adjust text entrance easing.
• [05:46] Claude executes Higgsfield MCP tool calls (ae_get_skill, ae_layer_info, etc.) to inspect and alter timeline keyframes and shape layers.
• [07:35] Presenter prompts Claude to generate three 3D isolated image assets for the intro words, remove their backgrounds via Higgsfield tools, and place them above text layers in the After Effects composition.
• [09:51] Final playback in After Effects displaying the updated colors, animation timings, and imported cutout graphics.

Claims & numbers

• The presenter states Anthropic released Claude Opus 5.5 on September 22, 2026.
• The presenter notes that Claude Opus 5.5 costs 40% less to run on default settings compared to Opus 5.
• On-screen documentation shows input and output token costs are $4 and $20 per million tokens (20% less than Opus 5), cache reads are $0.20 per million tokens (60% less), and it generates output over 30% faster than Opus 5.
• Official benchmark scores shown:
  • Agentic coding (Terminal-Bench 4.0): Opus 5.5 scores 66.4% (vs. 55.8% for Fable 5.1, 52.3% for Opus 5, 57.9% for GPT-6 Astra, 37.3% for GPT-5.6 Sol).
  • FrontierCode v1.1: Opus 5.5 scores 54.4% (vs. 50.3% for Fable 5.1, 48.0% for Opus 5, 53.3% for GPT-6 Astra).
  • Knowledge work (GDPval-AA v2.1): Opus 5.5 scores 1846 (vs. 1735 for Fable 5.1, 1708 for Opus 5).
  • Business workflows (AutomationBench): Opus 5.5 scores 40.0% (vs. 31.4% for Fable 5.1, 41.4% for GPT-6 Astra).
  • Multidisciplinary reasoning (Humanity's Last Exam): Opus 5.5 scores 67.7% with tools (vs. 65.6% for Fable 5.1, 63.6% for Opus 5).
  • Agentic scientific research (Terminal-Bench Science 0.1): Opus 5.5 scores 58.7% with tools (vs. 52.6% for Fable 5.1, 64.6% for GPT-6 Astra).
  • Computer use (OSWorld 2.0): Opus 5.5 achieves 81.8% partial score (vs. 80.7% for Fable 5.1, 74.0% for Opus 5).
  • Visual chart recognition (Chartography): Opus 5.5 scores 89.0%.
• On the Artificial Analysis Intelligence Index, Opus 5.5 is listed with an index score of 58.

Notable quotes

• [00:00] "Just last week Anthropic released Claude Opus 5.5."
• [02:02] "If you give them the proper tools, you'll be able to create these beautiful visualizations and be able to have it in a form where you can edit or you can hand it off to a designer."
• [06:23] "The cool thing with this is all of a sudden with these models is you really have the ability where you can control all of this directly from Claude Code."

Assessment
This is an authentic third-party product review and tutorial demonstrating real software integration. The presenter directly interacts with Claude and Adobe After Effects through a live Model Context Protocol (MCP) server, showing genuine execution logs, tool calls, and automated timeline adjustments without noticeable fabrication.

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