v1.19.0

What's Changed

• v1.19 by @willbakst in #868

Full Changelog: v1.18.4...v1.19.0

v1.18.4

What's Changed

• fix: missing reasoning_effort call param for openai by @willbakst in #844
• fix: call response tool message param typing issue by @willbakst in #843
• fix: incorrect exclusion of custom titles for fields in base models by @willbakst in #846
• tests: directly test GenerateJsonSchemaNoTitles by @willbakst in #847
• fix: code examples in WHY.md by @teamdandelion in #849
• chore: reorganize official sdk examples by @teamdandelion in #850
• feat: convert tools.md to provider-agnostic-calls by @teamdandelion in #848
• set max payload limit to 10mb for google because of base64 33% overhead by @CharlesOural in #838
• fix: mypy issue with litellm setup call by @willbakst in #856
• Fix 839 840 by @willbakst in #860
• fix: missing common Usage for llm module by @willbakst in #861
• feat: convert agents.md to provider agnostic calls by @teamdandelion in #857
• Docs updates 852 855 by @willbakst in #863
• feat: convert evals.md to provider-agnostic calls by @teamdandelion in #862
• fix: cost calculation should handle cached tokens by @willbakst in #865
• Fix 858 by @willbakst in #866
• Update ruff by @willbakst in #867

Full Changelog: v1.18.3...v1.18.4

v1.18.3

What's Changed

• pyright excludes .venv by @teamdandelion in #829
• feat: expose more imports through mirascope, core, and llm by @teamdandelion in #826
• feat: provider-agnostic Get Started docs by @teamdandelion in #825
• add google token count by @CharlesOural in #831
• Add google max payload size limit by @CharlesOural in #828
• Convert response_models.md to provider-agnostic calls by @teamdandelion in #830
• Convert json_mode docs to provider agnostic calls by @teamdandelion in #833
• convert output_parser to provider agnostic calls by @teamdandelion in #834
• fix: issue with response model validation + few-shot issues by @willbakst in #836
• fix: issue with llm.call async not having it's own inner async by @willbakst in #837

New Contributors

• @teamdandelion made their first contribution in #829

Full Changelog: v1.18.2...v1.18.3

v1.18.2

What's Changed

• Fix google image url upload part by @willbakst in #822

Full Changelog: v1.18.1...v1.18.2

v1.18.1

What's Changed

• Support latest azure sdk by @koxudaxi in #821

Full Changelog: v1.18.0...v1.18.1

v1.18.0

What's Changed

• feat: Add provider Google by @koxudaxi in #776
• docs: update tool examples field support by @koxudaxi in #818
• blog: apply bolding to code by @orlando21 in #799
• Deprecate gemini vertex by @willbakst in #819

Full Changelog: v1.17.0...v1.18.0

v1.17.0

What's Changed

• feat: fallback decorator by @willbakst in #814
• Fix llm.call issue and add support for URLs without downloading by @willbakst in #813
• feat: use system_instruction for gemini and vertex by @koxudaxi in #815
• Fallback docs by @willbakst in #816

Full Changelog: v1.16.9...v1.17.0

v1.16.9

What's Changed

• Fix issue with llm.call for agent loops with tools by @willbakst in #809

Full Changelog: v1.16.8...v1.16.9

v1.16.8

What's Changed

• blog: close parentheses by @orlando21 in #800
• blog: llm agents by @willbakst in #803
• Fix tooklit example, logfire support bump, remove extra computed field decorator by @willbakst in #807

Full Changelog: v1.16.7...v1.16.8

v1.16.7

What's Changed

• fix: issue with usage not being serializable by @willbakst in #798

Full Changelog: v1.16.6...v1.16.7