0.12.0 (2026-05-29)
- regenerate remaining packages (#17078) (cc4f36f65e8b29d28f636e082435801b899cbca0)
0.11.0 (2026-03-26)
- Allow Protobuf 7.x (1eb7c268482e55971966e284dac2cbeb903adcbb)
- update image to us-central1-docker.pkg.dev/cloud-sdk-librarian-prod/images-prod/python-librarian-generator@sha256:f5426423676c75008c2135037e7b98f78cbb99f78b3c46fe043b6897be92d836 (3654fe76d755dd8db62ece81d5770ec58b3624df)
- Require Python 3.9 (1eb7c268482e55971966e284dac2cbeb903adcbb)
- Require google-api-core >= 2.11.0 (1eb7c268482e55971966e284dac2cbeb903adcbb)
- Improve type checking (1eb7c268482e55971966e284dac2cbeb903adcbb)
0.10.0 (2026-01-08)
- add support for GoogleMaps and FileSearch tools (17cc12897e6afdf3f3131531c50a8226a3f57c0f)
- add support for RetrievalConfig (17cc12897e6afdf3f3131531c50a8226a3f57c0f)
- add IMAGE_RECITATION reason when generation is stopped due to recitation (17cc12897e6afdf3f3131531c50a8226a3f57c0f)
- add json_name for GenerationConfig.response_json_schema_ordered field (17cc12897e6afdf3f3131531c50a8226a3f57c0f)
- add support for GroundingMetadata.google_maps_widget_context_token (17cc12897e6afdf3f3131531c50a8226a3f57c0f)
- auto-enable mTLS when supported certificates are detected (c353aa5bcc937ef9399c8efc90492dadbcf01aa2)
- check Python and dependency versions in generated GAPICs (c353aa5bcc937ef9399c8efc90492dadbcf01aa2)
0.9.0 (2025-10-20)
- deprecation notice for HarmCategory.HARM_CATEGORY_CIVIC_INTEGRITY (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add support for additional reasons to stop image generation (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add support for finish_message that contain details the reason why the model stopped generating tokens (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add URL_RETRIEVAL_STATUS_UNSAFE and URL_RETRIEVAL_STATUS_PAYWALL statuses for url retrieval (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add log_probability_sum (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add Computer Use tool type (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add support for FunctionResponsePart (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add support for image generation features (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add support for additional reasons to stop image generation (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add support for raw media bytes for function response (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add log_probability_sum (84be9032eb7ba04bb305870160699d2bc12e0dac)
- add URL_RETRIEVAL_STATUS_UNSAFE and URL_RETRIEVAL_STATUS_PAYWALL statuses for url retrieval (84be9032eb7ba04bb305870160699d2bc12e0dac)
- replaced GenerateVideoResponse by PredictLongRunningGeneratedVideoResponse field (84be9032eb7ba04bb305870160699d2bc12e0dac)
0.8.0 (2025-10-16)
- Add support for Python 3.14 (98ee71abc0f97c88239b50bf0e0827df19630def)
- Deprecate credentials_file argument (98ee71abc0f97c88239b50bf0e0827df19630def)
0.7.0 (2025-08-29)
- add support for video_metadata
- add a flag whether the model supports thinking (4d08928)
- add a flag whether the model supports thinking (4d08928)
- add AudioTranscriptionConfig (4d08928)
- add MultiSpeakerVoiceConfig to speech config (4d08928)
- add support for function response scheduling in the conversation (4d08928)
- add support for thought_signature (4d08928)
- add support for UrlContextMetadata (4d08928)
- add support for video_metadata (4d08928)
- add time_range_filter to GoogleSearch tool type (4d08928)
- add UNEXPECTED_TOOL_CALL to the Candidate.FinishReason enum (4d08928)
- add url_context to GoogleSearch tool type (4d08928)
0.6.18 (2025-04-29)
- add CODE_RETRIEVAL_QUERY embedding task type (d530650)
- add CODE_RETRIEVAL_QUERY embedding task type (d530650)
- add language_code to speech config (d530650)
- add minimum, maximum, any_of, property_ordering, default and null type to Schema (d530650)
- add seed to GenerationConfig (d530650)
- add support for live API (googleapis/go-genai#275) (d530650)
- add support for ModalityTokenCounts in UsageMetadata responses (d530650)
- add support for ModalityTokenCounts in UsageMetadata responses (d530650)
- add support for Veo: PredictLongRunning and generated/downloadable files (d530650)
- add ThinkingConfig, MediaResolution, and seed to GenerationConfig (d530650)
0.6.17 (2025-03-15)
- [Many APIs] Allow Protobuf 6.x (a1b9294)
0.6.16 (2025-02-12)
- Add REST Interceptors which support reading metadata (a961bc0)
- Add support for reading selective GAPIC generation methods from service YAML (a961bc0)
0.6.15 (2025-01-13)
- Add BidiGenerateContent + all the necessary protos (2c1e359)
- Add civic_integrity toggle to generation_config (2c1e359)
- Add GoogleSearch tool type (2c1e359)
- Add id to FunctionCall and FunctionResponse (2c1e359)
- Add image_safety block_reason + finish_reason (2c1e359)
- Add response_modalities to generation_config (2c1e359)
- Add return type
Schema responseto function declarations (2c1e359) - Add TuningMultiturnExample (2c1e359)
- Add v1alpha (2c1e359)
- Add voice_config to generation_config (2c1e359)
- Update safety filter list to include civic_integrity (2c1e359)
0.6.14 (2024-12-12)
- Add support for opt-in debug logging (b1dfc55)
- Fix typing issue with gRPC metadata when key ends in -bin (b1dfc55)
0.6.13 (2024-11-21)
- Add GroundingMetadata.web_search_queries (3535963)
- Adds
GenerateContentResponse.model_versionoutput (3535963) - Adds search grounding (3535963)
- Some small updates. (3535963)
0.6.12 (2024-10-31)
- disable universe-domain validation (85c7512)
0.6.11 (2024-10-24)
0.6.10 (2024-09-23)
- Add GenerationConfig.{presence_penalty, frequency_penalty, logprobs, response_logprobs, logprobs} and Candidate.{avg_logprobs, logprobs_result} (d6238e4)
- Add GoogleSearchRetrieval tool and candidate.grounding_metadata (d6238e4)
- Add HarmBlockThreshold.OFF (d6238e4)
- Add HarmCategory.HARM_CATEGORY_CIVIC_INTEGRITY (d6238e4)
- Add PredictionService (for Imagen) (d6238e4)
- Add Schema.min_items (d6238e4)
- Add TunedModels.reader_project_numbers (d6238e4)
0.6.9 (2024-08-19)
- Add model max_temperature (fdebbf2)
- Add new PromptFeedback and FinishReason entries (fdebbf2)
- Add new PromptFeedback and FinishReason entries for google-gemini/deprecated-generative-ai-python#476 (fdebbf2)
- Many small fixes (fdebbf2)
0.6.8 (2024-07-30)
- Retry and timeout values do not propagate in requests during pagination (c6eeae0)
0.6.7 (2024-07-08)
0.6.6 (2024-06-26)
0.6.5 (2024-06-11)
- Add cached_content_token_count to CountTokensResponse (09c7fae)
- Add cached_content_token_count to generative_service's UsageMetadata (09c7fae)
- Add content caching (09c7fae)
- Small fixes (09c7fae)
0.6.4 (2024-05-16)
- v1: Add generate_content_request to CountTokensRequest (e5dd7ed)
- v1: Add usage metadata to GenerateContentResponse (e5dd7ed)
- v1beta: Add video metadata to files API (e5dd7ed)
- v1beta: Update timeouts for generate content (e5dd7ed)
- v1: Update timeouts (e5dd7ed)
0.6.3 (2024-05-07)
0.6.2 (2024-04-15)
- v1: Add
output dimensionalitytoEmbedContentRequestmessage (e3bfbc6) - v1: Add
QUESTION_ANSWERINGandFACT_VERIFICATIONtoTaskTypeEnum (e3bfbc6) - v1: Add rest binding for tuned models (e3bfbc6)
- v1beta: Add
output dimensionalitytoEmbedContentRequestmessage (e2cf0c4) - v1beta: Add
QUESTION_ANSWERINGandFACT_VERIFICATIONtoTaskTypeEnum (e2cf0c4) - v1beta: Add
response_mime_typetoGenerationConfigmessage (e2cf0c4)
0.6.1 (2024-04-03)
- Add file upload service (c0a0bf6)
- Add file_data to part options (c0a0bf6)
- Add system instructions (c0a0bf6)
- Add tool_config (c0a0bf6)
- A comment for field
candidate_countin message.[google.ai](https://www.google.com/url?sa=D&q=http%3A%2F%2Fgoogle.ai).generativelanguage.v1beta.GenerationConfigis changed (c0a0bf6)
0.6.0 (2024-03-22)
- make learning rate a one-of
- Add
learning_rate_multiplierto tuningHyperparameters(35017ea)
- make learning rate a one-of (35017ea)
- A few small updates (35017ea)
0.5.4 (2024-03-05)
0.5.3 (2024-02-22)
- [Many APIs] fix
ValueErrorintest__validate_universe_domain(#12280) (2d75d0e) - deps: [Many APIs] Require
google-api-core>=1.34.1(#12303) (fbb80c3)
0.5.2 (2024-02-06)
- Add google-auth as a direct dependency (780c5f1)
- Add staticmethod decorator to _get_client_cert_source and _get_api_endpoint (780c5f1)
- Resolve AttributeError 'Credentials' object has no attribute 'universe_domain' (780c5f1)
0.5.1 (2024-02-01)
- Allow users to explicitly configure universe domain (a0392ee)
0.5.0 (2024-01-24)
- Fix content.proto's Schema -
typeshould be required
- Update GenAI libraries to include input_safety_feedback (d2004d4)
- Fix content.proto's Schema -
typeshould be required (d2004d4)
- Minor docs updates (d2004d4)
- Update summary, improve description for
titleinEmbedContentRequest(d2004d4)
0.4.1 (2024-01-22)
- [google-ai-generativelanguage] Fixed minor documentation typos for field
function_declarationsin messagegoogle.ai.generativelanguage.v1beta.Tool(#12206) (52957f3)
0.4.0 (2023-12-09)
- Add v1, contains only GenerativeService, nothing else (23d8814)
- Add v1beta, adds GenerativeService and RetrievalService (23d8814)
- Set
google.ai.generativelanguage_v1betaas the default import (23d8814)
0.3.5 (2023-12-07)
- Add support for python 3.12 (ea6cfc2)
- Introduce compatibility with native namespace packages (ea6cfc2)
0.3.4 (2023-09-30)
0.3.3 (2023-09-21)
0.3.2 (2023-09-20)
- Add BatchEmbedText and CountTextTokens to the text service (38f2ca3)
- Add google/ai/generativelanguage_v1beta3 (38f2ca3)
- Add model tuning (38f2ca3)
- Add permissions service (38f2ca3)
0.3.1 (2023-07-05)
0.3.0 (2023-06-29)
- remove
BLOCK_NONEfromHarmBlockThreshold(9d46f7f)