Skip to main content
llm.info

Claude 3 Opus

Anthropic

Anthropic's most capable model in the Claude 3 family, setting new standards for AI performance. Outperforms peers on most common evaluation benchmarks including graduate-level reasoning (GPQA), undergraduate knowledge (MMLU), and basic math (GSM8K). Achieved near-perfect 99% accuracy on 'Needle in a Haystack' recall evaluation. Features sophisticated vision capabilities for analyzing photos, charts, and technical diagrams. 200K context window expandable to 1M for specific use cases. Note: Deprecated as of January 2026, scheduled for retirement January 5, 2026 - migrate to Claude Opus 4.1.

Strengths

  • Near-perfect 99% accuracy on Needle in a Haystack recall evaluation
  • Outperforms GPT-4 and Gemini Ultra on most common benchmarks
  • Sophisticated vision capabilities - analyzes photos, charts, graphs, technical diagrams
  • 200K context window with option to expand to 1M for specific use cases
  • Graduate-level expert reasoning (GPQA benchmark leader)
  • Strong function calling and tool use for agentic workflows
  • Available via Claude.ai Pro for $20/month unlimited access

Caveats

  • Deprecated - scheduled for full retirement January 5, 2026 (migrate to Opus 4.1)
  • Expensive pricing ($15 input, $75 output vs GPT-4o $2.50/$10)
  • August 2023 knowledge cutoff older than competitors
  • 4K max output tokens lower than newer models
  • Slower generation than Claude 3 Haiku and Sonnet
Vendor
Anthropic
Context Window
200,000 tokens
Max Output
4,096 tokens
License
Proprietary
Release Date
2024-03-04
Training Cutoff
2023-08-01
Modalities
text
image
BenchmarkScoreMax
livebench_instruction_following65.6100.0
livebench_language52.6100.0
livebench_coding38.6100.0

Capabilities

Vision
Audio
Video
Tool Use

Pricing

$15 per 1M input tokens, $75 per 1M output tokens. ChatGPT Pro: $20/month unlimited access

Reviews

Comments