No longer supported

This model is legacy (released November 2023) and is no longer actively maintained or recommended by Anthropic. Consider using their current flagship model instead.

legacylegacyllm Proprietary

Claude 2.1

by Anthropic· Released November 2023· Cutoff Early 2023

Claude 2.1 is an updated version of Claude 2, featuring a doubled context window of 200K tokens and reduced hallucination rates. It offers improved accuracy, lower rates of false refusals, and system prompt support, making it suitable for complex reasoning and long-document analysis.

Official Site API Docs

Input cost

$8.00 per 1M tokens

Output cost

$24.00 per 1M tokens

Context window

200K tokens

Max output

4096 tokens

Modalities

text

License

proprietary

Capabilities

Long ContextSystem PromptJSON ModeStreamingFunction CallingCode GenerationMultilingual

Best For

Complex reasoning tasks requiring very long context windows, such as analyzing entire codebases or lengthy documents.

Strengths

200K token context window
Reduced hallucination rates compared to Claude 2
Lower rates of false refusals
Strong at nuanced reasoning and analysis

Limitations

No vision or image input
Legacy model, superseded by Claude 3 family
Higher latency than newer models
No multimodal capabilities

Use Cases

Long-document summarization

Codebase analysis and debugging

Legal contract review

Research paper analysis

Complex question answering over large texts

Data extraction from lengthy reports

Conversational agents with long-term memory

Improvements Over Previous Model

Context window doubled from 100K to 200K tokens
Reduced hallucination rates by 2x compared to Claude 2
Lower rates of false refusals
Improved accuracy on complex reasoning tasks
System prompt support added

Back to all models