Claude 1.3

Claude 1.3 is a large language model by Anthropic released on April 18, 2023.

Compared to Claude 1.2, Anthropic described the model as “more robust against red-team inputs, better at precise instruction-following, better at code, and better and non-English dialogue and writing.”¹ Support for 100k context (up from 9k) was introduced on May 11, 2023.²

Claude 1.3’s pretraining data cutoff was around late 2022 or early 2023.¹

In July 2023, Anthropic published evaluations comparing Claude 1.3, Instant 1.1, Claude 2.0, and Claude 1.3 Helpful-only.¹

An LLM with the same base model as Claude 1.3 was used in the paper “Question Decomposition Improves the Faithfulness of Model-Generated Reasoning” (Radhakrishnan et al., Jul 17, 2023).