0.15.1

• Now forwards URLs directly to Anthropic's API for images and PDFs, if a URL has been provided. #24

0.15

• Support for schemas using the new llm prompt --schema X and model.prompt(..., schema=X) mechanisms. #22

0.15a0

• Support for schemas using the new llm prompt --schema X and model.prompt(..., schema=X) mechanisms. #22

0.14.1

• Pin to anthropic>=0.47.2 to avoid Messages.stream() got an unexpected keyword argument 'thinking' error. #20

0.14

See also these annotated release notes.

• Support for the new Claude 3.7 Sonnet model, including -o thinking 1 to turn on extended thinking mode and -o thinking_budget X to increase the thinking token budget from the 1024 default. #14
• Claude 3.5 Haiku now supports image inputs. #17
• Fixed a bug that occurred when continuing an existing conversation using --async mode. #13
• Fixed a bug where max_tokens and temperature were logged in the database even when using their default options. #16

0.13

• Now depends on LLM 0.22, adding support for the new model.prompt(..., key="key") parameter. #12
• README now features nicer aliases for the models. Thanks, Bill Mill. #11

0.12

• Support for Claude's prefill feature, using the new -o prefill '{' option and the accompanying -o hide_prefill 1 option to prevent the prefill from being included in the output text. #2
• New -o stop_sequences '```' option for specifying one or more stop sequences. To specify multiple stop sequences pass a JSON array of strings :-o stop_sequences '["end", "stop"].
• Model options are now documented in the README.

0.11

• Renamed llm-claude-3 to llm-anthropic
• Now looks for a key called anthropic (previously looked for claude) - set that with llm keys set anthropic
• Model IDs now start with a anthropic/ prefix, e.g. llm -m anthropic/claude-3-5-sonnet-latest 'say hi' - but previous aliases such as llm -m claude-3-5-sonnet 'say hi' continue to work.