Commit Graph

  • cf5f5f41f4 feat(llms): enhance structured output handling and validation feat/adaptive-thinking Dmitry Ng 2026-07-17 12:34:33 +03:00
  • 1fc129969f feat(llms): add provider-neutral structured output via WithStructuredOutput Dmitry Ng 2026-07-17 04:39:05 +03:00
  • 3079a9e746 feat(anthropic): surface model refusals as a typed error Sergey Kozyrenko 2026-07-13 20:28:06 +07:00
  • c1eca2bae7 feat(googleai): map reasoning effort to Gemini 3.x thinking_level Sergey Kozyrenko 2026-07-13 20:28:04 +07:00
  • d9b6a77002 feat(bedrock): honor explicit reasoning-off on Converse and legacy paths Sergey Kozyrenko 2026-07-13 16:15:57 +07:00
  • 42e621418c feat(googleai): honor explicit reasoning-off via zero thinking budget Sergey Kozyrenko 2026-07-13 16:15:40 +07:00
  • 31119ac596 feat(openai): honor explicit reasoning-off via reasoning_effort none Sergey Kozyrenko 2026-07-13 16:15:27 +07:00
  • 52ee3b16ad feat(anthropic): honor explicit reasoning-off via the disable wire Sergey Kozyrenko 2026-07-13 16:15:14 +07:00
  • 904fa6c8df feat(llms): add ReasoningSupport capability introspection Sergey Kozyrenko 2026-07-13 16:14:55 +07:00
  • f6774dad78 feat(reasoning): add model-aware reasoning-off resolver (ResolveOff) Sergey Kozyrenko 2026-07-13 16:14:39 +07:00
  • 064db9d0c1 feat(llms): add tri-state ReasoningMode with backward-compatible resolver Sergey Kozyrenko 2026-07-13 16:14:20 +07:00
  • e9bf4bb222 fix(bedrock): pin temperature and drop top_p/top_k with budget thinking Sergey Kozyrenko 2026-07-13 12:34:07 +07:00
  • 6d4febb58f fix(openai): keep tool-call argument bytes when the name is uncached Sergey Kozyrenko 2026-07-13 12:10:12 +07:00
  • 39fde658d6 fix(llms): default adaptive reasoning with no effort to high in one place Sergey Kozyrenko 2026-07-13 12:04:03 +07:00
  • 886d678046 feat(llms): resolve Claude thinking mechanism from the model Sergey Kozyrenko 2026-07-13 11:51:10 +07:00
  • e7e7305766 test(llms): make reasoning test comments self-contained Sergey Kozyrenko 2026-07-13 10:58:16 +07:00
  • 8dde9db229 fix(anthropic): drop top_p with budget thinking to satisfy the API Sergey Kozyrenko 2026-07-13 10:58:14 +07:00
  • 49510f5d85 chore(llms): refresh default models to current generations Sergey Kozyrenko 2026-07-13 02:22:08 +07:00
  • c56a5c911f feat(bedrock): add model constants for the mid-2026 catalog Sergey Kozyrenko 2026-07-13 02:22:07 +07:00
  • f1b7e01de1 docs(llms): correct the adaptive-thinking model range in option docs Sergey Kozyrenko 2026-07-13 01:48:36 +07:00
  • 2a55177d3e fix(anthropic): send the interleaved-thinking beta only for budget thinking Sergey Kozyrenko 2026-07-13 01:48:36 +07:00
  • 868cd14b2a fix(bedrock): keep Converse adaptive thinking for newer Claude families Sergey Kozyrenko 2026-07-13 01:48:09 +07:00
  • 8314dfedb0 fix(bedrock): honor adaptive thinking on the legacy InvokeModel path Sergey Kozyrenko 2026-07-13 01:47:47 +07:00
  • 3c3a32e1ff test(openai): assert xhigh/max reach the request body Sergey Kozyrenko 2026-06-26 04:26:06 +07:00
  • 12b63aaacd fix(llms): remove stale xhigh/max clamp in GetEffort Sergey Kozyrenko 2026-06-26 03:08:35 +07:00
  • c85de7639e Merge origin/feat/adaptive-thinking: keep audited adaptive-thinking work Sergey Kozyrenko 2026-06-21 22:34:09 +07:00
  • 66e1d3f14f feat: native adaptive thinking for anthropic and bedrock Sergey Kozyrenko 2026-06-21 20:19:59 +07:00
  • 91d8dfdd9c feat: native adaptive thinking for anthropic and bedrock Sergey Kozyrenko 2026-06-21 20:19:59 +07:00
  • 2a7e660e97 chore: upgrade ollama dependency to v0.18.0 across all modules main-vxcontrol v0.1.14-update.5 release/v0.1.14-update.5 Dmitry Ng 2026-03-15 18:58:21 +03:00
  • 5835a5ab5a fix: cache tool call names in streaming to handle providers with incomplete chunks Dmitry Ng 2026-03-15 18:48:26 +03:00
  • 649d2f5b67 fix: normalize multiple tool calls in Bedrock Converse API Dmitry Ng 2026-03-15 16:54:19 +03:00
  • 1b1c761d98 chore: remove unmaintained legacy LLM providers v0.1.14-update.4 release/v0.1.14-update.4 Dmitry Ng 2026-03-11 18:55:16 +03:00
  • 560049b5f2 feat: implement extra body support in chat requests and enhance testing Dmitry Ng 2026-03-11 17:32:18 +03:00
  • f2c57c4720 feat: breaking changes with llms.CallOptions and improve test coverage v0.1.14-update.3 release/v0.1.14-update.3 Dmitry Ng 2026-03-10 19:25:27 +03:00
  • 274ec52356 refactor: streamline test case structure and enhance message handling v0.1.14-update.2 release/v0.1.14-update.2 Dmitry Ng 2026-03-04 03:22:36 +03:00
  • 8d98658ae2 feat: upgrade ollama dependency to v0.17.5 and implement cloud support Dmitry Ng 2026-03-04 02:52:56 +03:00
  • 1c3dd0fca4 feat: update AWS SDK dependencies and enhance caching support Dmitry Ng 2026-03-04 01:08:38 +03:00
  • a17a86c171 chore: update golangci-lint version and refactor test cases v0.1.14-update.1 release/v0.1.14-update.1 Dmitry Ng 2026-02-22 21:31:38 +03:00
  • fed202eccd fix: update Node.js version in publish-docs workflow Dmitry Ng 2026-02-22 18:44:34 +03:00
  • 9b82b86a11 chore: update React and related dependencies to latest versions Dmitry Ng 2026-02-22 18:35:14 +03:00
  • c150dc39fe chore: update dependencies for ollama to latest versions Dmitry Ng 2026-02-22 17:40:06 +03:00
  • ef220222bd llms: standardize JSON keys in ContentReasoning struct and add comprehensive tests Dmitry Ng 2026-02-03 13:13:54 +03:00
  • 7c01254437 llms: enhance token usage mapping and cost calculation for multiple providers Dmitry Ng 2026-02-03 12:30:48 +03:00
  • b1ffb5ff9b llms/googleai: enhance handling of empty text parts in convertParts function Dmitry Ng 2026-02-02 22:44:14 +03:00
  • 712d720afd llms: add reasoning support for message content serialization Dmitry Ng 2026-02-02 21:57:04 +03:00
  • 6eb6b7ca8c llms/openai: add reasoning content preservation for multi-turn tool calls Dmitry Ng 2026-02-02 15:45:02 +03:00
  • 2109b1332d httputil: enhance ApiKeyTransport with improved path handling and security Dmitry Ng 2026-01-29 19:10:15 +03:00
  • a41e95bdd0 httputil: enhance ApiKeyTransport with BaseURL and ProxyURL support Dmitry Ng 2026-01-29 15:08:38 +03:00
  • bb226cf639 llms/googleai: add comprehensive tests for signature handling in Gemini LLM and fix specific case only for gemini provider Dmitry Ng 2026-01-28 19:53:31 +03:00
  • f79b68a0db llms/bedrock: standardize token field names for cross-provider compatibility Dmitry Ng 2026-01-28 18:05:51 +03:00
  • 63e01f2f31 test: refactor and enhance caching tests for Anthropic LLM Dmitry Ng 2026-01-28 12:43:21 +03:00
  • fc8a08a23f test: add comprehensive tests for OpenAI LLM functionality Dmitry Ng 2026-01-28 12:42:27 +03:00
  • b86614de61 fix: update reasoning model checks to include additional identifiers Dmitry Ng 2026-01-28 12:41:43 +03:00
  • 59b196e98a feat: enhance reasoning and caching support across various LLMs Dmitry Ng 2026-01-27 22:56:49 +03:00
  • fc989b35d1 httputil: Implement circular buffer for identical requests in httprr Dmitry Ng 2026-01-26 13:42:49 +03:00
  • aa2dd71269 Merge branch 'main' into release/v0.1.14-update.1 Dmitry Ng 2026-01-24 02:07:00 +03:00
  • 8fea3de636 llms/openai: add web search tool support main lif 2026-01-12 00:59:56 +08:00
  • 509308ff01 agents: allow match from multiple lines for parseOutput function (#1415) v0.1.14 hemarina 2025-10-19 22:14:29 -07:00
  • 016539d88a llms/openai: sanitize HTTP errors to prevent API key exposure (#1419) Travis Cline 2025-10-19 17:03:32 -07:00
  • db2a947ba4 agents: fix ChainCallOption silent failure (#1420) Travis Cline 2025-10-19 17:03:11 -07:00
  • 8e8a5408df anthropic: add improved streaming thinking/reasoning token support (#1418) Travis Cline 2025-10-15 15:22:49 -07:00
  • 334f4c3689 vectorstores/milvus: complete migration to new SDK v2 client (#1403) Travis Cline 2025-10-15 14:45:48 -07:00
  • 900c1b9498 httputil: enhance JSONDebugClient with SSE and header debugging (#1404) Travis Cline 2025-09-16 23:34:52 -07:00
  • 1bbaab130f documentloaders: add RecursiveDirectoryLoader that loads documents with allowed extensions from a directory (#1386) zhu733756 2025-09-15 10:18:12 +08:00
  • 8c8b5e6084 llms: update model context sizes for GPT-4o and GPT-4 Turbo (#1389) Paul Negedu 2025-09-14 21:17:55 -05:00
  • 95afeb6e4e examples: fix typo in openai-chat-example.go of maximum tokens per generation, old: 104 tokens, fix: 1024 tokens (#1392) Alexander Nalobin 2025-09-15 05:17:23 +03:00
  • 601dc553d5 docs: fix API references and complete integration documentation (#1402) Travis Cline 2025-09-14 23:37:03 +02:00
  • 450c02a722 examples: update to v0.1.14-pre.4 (#1401) Travis Cline 2025-09-14 20:54:54 +02:00
  • dff679c481 docs: complete the incomplete basic chat app tutorial (#1399) Travis Cline 2025-09-14 20:20:59 +02:00
  • 0cc366260b vectorstores/milvus: update deprecation notice for archived SDK (#1398) Travis Cline 2025-09-14 20:20:44 +02:00
  • 33e7f8f55d tools/scraper: add max pages limit for scraper functionality HJJ 2025-09-09 01:28:30 +08:00
  • 91fc754372 llms: add prompt caching and reasoning token support (#1394) v0.1.14-pre.4 Travis Cline 2025-09-08 18:44:38 +02:00
  • bc181f1f52 agents,llms/anthropic: prevent panics in agent parsing and Anthropic responses (#1380) Travis Cline 2025-08-29 18:13:35 +02:00
  • 6d42f059f1 tools/scraper: fix context cancellation hanging (#1383) Travis Cline 2025-08-29 15:49:39 +02:00
  • 480112bee0 internal/httprr, vectorstores/azureaisearch: fix race conditions in tests (#1381) Travis Cline 2025-08-29 15:48:17 +02:00
  • 7cbb9b05f7 examples/groq: fixup groq example (#1385) Travis Cline 2025-08-29 15:48:02 +02:00
  • f1213d92d8 examples/groq: fix max_tokens compatibility issue (#1382) Travis Cline 2025-08-29 15:39:07 +02:00
  • a3fb67edc8 Merge pull request #1376 from kid-icarus/rk/filter-openai-metadata Travis Cline 2025-08-24 17:45:36 +02:00
  • 5ce2bab8a5 Merge pull request #1278 from rainu/openai-expose-token-details Travis Cline 2025-08-24 17:33:54 +02:00
  • f1eed5e215 Merge pull request #1379 from tmc/update-examples-to-v0.1.14-pre.3 Travis Cline 2025-08-23 19:23:12 +02:00
  • 827a36f321 examples: update examples to v0.1.14-pre.3 Travis Cline 2025-08-23 19:17:47 +02:00
  • 6c217b3865 test: improve test coverage and optimize storage with gzip compression (#1378) v0.1.14-pre.3 Travis Cline 2025-08-23 19:06:20 +02:00
  • dd61fd90f4 test: improve test resilience and update provider APIs (#1377) Travis Cline 2025-08-22 18:13:13 +02:00
  • 2f1f573d1c fix: filter out open ai max token metadata Ryan Kois 2025-08-20 11:00:23 -07:00
  • 0ddbdfda92 openai: fix temperature handling for reasoning models (GPT-5, o1, o3) (#1374) Travis Cline 2025-08-20 19:04:40 +02:00
  • fd690c56ea Tools Doc : URL was not working. (#1309) Zaber Al Rafian 2025-08-20 00:37:25 +06:00
  • 6d75f33614 examples: update examples to v0.1.14-pre.2 (#1373) Travis Cline 2025-08-19 20:36:29 +02:00
  • 9fc3d0aa27 llms/ollama: fix panic when context is cancelled during streaming (#1372) Travis Cline 2025-08-19 19:35:18 +02:00
  • fd8379f966 openai: add robust max_tokens support with backward compatibility (#1371) v0.1.14-pre.2 Travis Cline 2025-08-19 18:45:18 +02:00
  • ea5209cc18 googleai: fix user provided embedding model getting overridden Antonio Calabrese 2025-08-19 17:33:46 +01:00
  • 78296871ac mongovector: fix expected error string in non-tokenized filter test Matt Dale 2025-08-19 09:33:08 -07:00
  • 00ccb89a62 Revert "openai: add support for max_tokens (#1359)" (#1369) kid-icarus 2025-08-18 10:46:44 -07:00
  • 99951bca92 httprr: improve test recording stability and add utilities (#1368) Travis Cline 2025-08-18 19:41:52 +02:00
  • f94b2cc876 llms/bedrock: complete Anthropic tool calling support (#1367) Travis Cline 2025-08-18 17:17:05 +02:00
  • 182be016f8 bedrock: add tool calling support for Anthropic Claude models (#1327) Manish 2025-08-18 19:23:29 +05:30
  • 2a139b1e0d fix: add ReasoningContent field to openai GenerateContent functio… (#1324) topjohncian 2025-08-18 21:53:19 +08:00
  • 168c3e58d1 fix: ensure googleai schema conversion works with nested objects and arrays (#1326) kid-icarus 2025-08-18 06:53:09 -07:00
  • 76ce69b128 llms/bedrock: add modelProvider option and Nova model support (#1346) Dreamans 2025-08-18 20:03:37 +08:00
  • da67c4b88b Anthropic: allow streaming responses with tool use (#1343) James Pozdena 2025-08-18 06:02:57 -06:00
  • 6e7fe3ea3e openaiclient: add dimensions option while creating embeddings (#1338) Jeremy L. 2025-08-18 14:02:17 +02:00