Claude 2.1

Claude 2.1

Anthropic · Released Nov 21, 2023 Emerging
Intelligence #470 / 583
20.9 Our Score
AA Index #331 / 382
3.9 Artificial Analysis
Input
— Not priced
Output
— Not priced
Context
— Not reported

Analysis Summary

Claude 2.1 is an older Anthropic model released in November 2023. With an intelligence index of 3.9, GPQA of 0.319, and LiveCodeBench of 0.195, it sits well below current-generation models across all key capability dimensions. It has been superseded by Claude 3, 3.5, 3.7, and the Claude 4 family.

For business use, it is not recommended for new deployments. Reasoning, coding, and instruction-following performance are all materially weaker than what is available from Anthropic and other providers today. It lacks agentic index data, further limiting its suitability for tool-use or autonomous workflows.

Teams still running Claude 2.1 in production should prioritise migration to a current-generation model. Even mid-tier current models from Mistral or Meta offer substantially better performance at comparable or lower cost.

Assessed June 17, 2026

Editorial notes

Claude 2.1 from Anthropic is a significantly older model with limited benchmark performance across reasoning, coding, and instruction following; superseded by multiple Claude generations and not recommended for new deployments.

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

Performance Profile

Intelligence1.3Technical2.1Value0Content3.5
Intelligence 1.3/10
Technical 2.1/10
Content 3.5/10
Value 0/10

How Claude 2.1 compares

Claude 2.1 ranks #331 of 382 AI models we track for overall intelligence, #91 of 111 for coding. Claude 2.1 is currently free to use via OpenRouter.

Performance Indices

Source: Artificial Analysis

3.9 Intelligence Index
14 Coding Index

Benchmark Scores

Intelligence

GPQA Diamond 31.9% Graduate-level scientific reasoning
HLE 4.2% Humanity's Last Exam
MMLU Pro 49.5% Multi-task language understanding
MATH 500 37.4% Mathematical problem-solving
AIME 3.3% Competition mathematics
SciCode 18.4% Scientific computing

Technical

LiveCodeBench 19.5% Live coding evaluation

Benchmark data from Artificial Analysis and Hugging Face

How does Claude 2.1 stack up?

Compare side-by-side with other emerging models.

Compare Models

Model Information

ProviderAnthropic
Release Date November 21, 2023
Status Active

Leaderboard Categories

Frequently asked questions about Claude 2.1

How much does Claude 2.1 cost?

Claude 2.1 is currently available for free via OpenRouter.

Is Claude 2.1 good for coding?

On our coding benchmark index, Claude 2.1 ranks #91 of 111 models, placing it in the broader range of the field for code generation and debugging.

Who created Claude 2.1?

Claude 2.1 is developed by Anthropic and was released on November 21, 2023.