3 Releases published by 1 person

• v1.74.15.rc.2

  published Aug 5, 2025

• v1.75.0-nightly

  published Aug 5, 2025

• v1.75.0.dev2

  published Aug 5, 2025

24 Pull requests merged by 5 people

• Exclude none fields on /chat/completion - fixes n8n bug + Allow calling /v1/models when end user over budget

  #13320 merged Aug 6, 2025

• feat(JinaAI): support multimodal embedding models

  #13181 merged Aug 6, 2025

• fix(streaming_handler.py): include cost in streaming usage object

  #13319 merged Aug 6, 2025

• [LLM Translation] Fix model group on clientside auth with API calls

  #13314 merged Aug 6, 2025

• [Bug Fix] [Bug]: New Databricks Foundation Models databricks-gpt-oss-20b and databricks-gpt-oss-120b failed with error: litellm.APIConnectionError: 'signature'

  #13318 merged Aug 6, 2025

• [Feat] - When using custom tags on prometheus allow using wildcard patterns

  #13316 merged Aug 6, 2025

• [Bug]: Fix Mimetype Resolution Error in Bedrock Document Understanding

  #13309 merged Aug 6, 2025

• Litellm fix OpenAI spec tools

  #13315 merged Aug 6, 2025

• [MCP Gateway] refactor mcp guardrails

  #13238 merged Aug 5, 2025

• Fix double slash issue in SSO login URL construction

  #13289 merged Aug 5, 2025

• [Proxy server] Add apscheduler log suppress

  #13299 merged Aug 5, 2025

• [Redis IAM] Change documentation

  #13306 merged Aug 5, 2025

• [LLM Translation] Fix model group on clientside auth with API calls

  #13293 merged Aug 5, 2025

• Create New Key - Make Team Field Required for Service Account

  #13302 merged Aug 5, 2025

• [Feat] Add fireworks gpt-oss models

  #13303 merged Aug 5, 2025

• New models - add fireworks_ai/glm-4p5 model family

  #13297 merged Aug 5, 2025

• [New model] add bedrock/us.anthropic.claude-opus-4-1-20250805-v1:0

  #13295 merged Aug 5, 2025

• [LLM Translation] claude opus 4.1 support for anthropic provider

  #13296 merged Aug 5, 2025

• [LLM Translation + Coding tools] Added litellm claude code count tokens support

  #13261 merged Aug 5, 2025

• [Redis] - Add ability to add client through GCP IAM Auth

  #13275 merged Aug 5, 2025

• Revert "Fix: Langfuse reporting "client closed" error due to httpx client TTL"

  #13291 merged Aug 5, 2025

• fix(main.py): handle tool being a pydantic object + Fix unpack defs deepcopy issue for bedrock

  #13274 merged Aug 5, 2025

• Ensure disable_llm_api_endpoints works + Add wildcard model support for 'team-byok' model

  #13278 merged Aug 5, 2025

• fix OCI linting errors

  #13279 merged Aug 5, 2025

5 Pull requests opened by 5 people

• Fix/create vector store

  #13285 opened Aug 5, 2025

• 🆕 [FEATURE]: Add WandB by Coreweave Inference Endpoints as a hub.

  #13290 opened Aug 5, 2025

• Add Custom Tooltips to Model Mapping Table

  #13294 opened Aug 5, 2025

• Feat/sambanova embeddings

  #13308 opened Aug 5, 2025

• [MCP Gateway] fix auth on ui for bearer servers

  #13312 opened Aug 5, 2025

25 Issues closed by 4 people

• [Bug]: Listing models should work even when end-user budget is exceeded

  #13286 closed Aug 6, 2025

• [Bug]: tool_calls coming back as null while they should be Array

  #13055 closed Aug 6, 2025

• [Feature]: Return x-litellm-response-cost header when streaming with include_usage: true

  #12689 closed Aug 6, 2025

• [Bug]: New Databricks Foundation Models databricks-gpt-oss-20b and databricks-gpt-oss-120b failed with error: litellm.APIConnectionError: 'signature'

  #13313 closed Aug 6, 2025

• [Bug]: Mimetype Resolution Error in Bedrock Document Understanding

  #12260 closed Aug 6, 2025

• [Feature]: `function_to_dict` supporting type unions

  #4249 closed Aug 6, 2025

• [Feature]: `function_to_dict` supporting defaulted arguments

  #4250 closed Aug 6, 2025

• [Bug]: inability to use `DeploymentTypedDict` in Pydantic `TypeAdapter` with Python<3.12

  #5664 closed Aug 6, 2025

• [Bug]: Router not respecting TPM limits in concurrent async calls

  #5783 closed Aug 6, 2025

• [Bug]: randomly logs blow up with `router_cooldown_event_callback but _deployment is None for deployment_id=abc123. Doing nothing`

  #5933 closed Aug 6, 2025

• [Feature]: convenience `Enum` for `tool_choice`

  #6091 closed Aug 6, 2025

• [Bug]: `litellm.acompletion` with `stream`ing is unexpectedly manipulating the input `model_list[0]["litellm_params"]`

  #6156 closed Aug 6, 2025

• [Bug]: `litellm.text_completion` not respecting `model_list`

  #6157 closed Aug 6, 2025

• [Feature]: automated nesting when using litellm sdk within langfuse observe() decorated function

  #8423 closed Aug 6, 2025

• Google Authentication Issue

  #8424 closed Aug 6, 2025

• API Keys not displaying after creation

  #8446 closed Aug 6, 2025

• [Feature]: Add Qwen2.5-3B-Instrcut

  #10437 closed Aug 6, 2025

• [Bug]: Azure Audio Transcription & DALL-E (Access denied)

  #10440 closed Aug 6, 2025

• [Bug]: Random problems with docker images using OpenSSL 3.5

  #10444 closed Aug 6, 2025

• [Bug]: gemini request body is not logged

  #10449 closed Aug 6, 2025

• [Feature]: Add "claude-opus-4-1-20250805" in "model_prices_and_context_window.json"

  #13305 closed Aug 5, 2025

• [Bug]: Claude Code count_tokens API is not implemented in the LiteLLM proxy.

  #13252 closed Aug 5, 2025

• [Bug]: LiteLLM Gateway crashes when processing request to Bedrock with tool of JSON Schema that contains circular $ref references

  #13151 closed Aug 5, 2025

• [Bug]: Tool calling broken from expecting "keys" on OpenAITool (v1.74.9)

  #13064 closed Aug 5, 2025

• [Bug]: api request works even if DISABLE_LLM_API_ENDPOINTS=true

  #13095 closed Aug 5, 2025

16 Issues opened by 16 people

• [Bug]: Add New MCP Server : Connection Status shows Connection Failed - Failed to fetch

  #13322 opened Aug 6, 2025

• [Bug]: LiteLLM Bedrock Guardrail Error Response Incomplete

  #13321 opened Aug 6, 2025

• [Feature]: Support for routing requests to the Cohere `v2/chat` api when submitted through OpenAI SDK

  #13311 opened Aug 5, 2025

• [Bug]: Model Alias Resolution Causes Permission Check Failure

  #13310 opened Aug 5, 2025

• [Feature]: SambanNova embeddings

  #13307 opened Aug 5, 2025

• [Bug]: DBRX - Signature Error with OSS GPT models

  #13304 opened Aug 5, 2025

• [Bug]: max_tokens parameter causes BedrockException “max_tokens_to_sample: Extra inputs are not permitted” for Claude Opus 4.1 model

  #13301 opened Aug 5, 2025

• [Feature]: Support reasoning in harmony response format for gpt-oss models

  #13300 opened Aug 5, 2025

• [Bug]: Empty list of MCP tools returned on streamable connection

  #13298 opened Aug 5, 2025

• litellm whether apikey in config file can be used for client authentication without connecting to the database.

  #13292 opened Aug 5, 2025

• [Bug]: LiteLLM_MCPServerTable.server_name does not exist (missing migration?)

  #13288 opened Aug 5, 2025

• [Bug]: Create Vector Store fail

  #13284 opened Aug 5, 2025

• [Bug]: APIConnectionError - negative file descriptor with Claude models

  #13283 opened Aug 5, 2025

• [Bug]: Migration using non-root image failed

  #13282 opened Aug 5, 2025

• Changes to the Bitnami Chart and Image Catalog

  #13281 opened Aug 5, 2025

• [Bug]: Missing spend logs

  #13280 opened Aug 5, 2025

41 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Ensure that `function_call_prompt` extends system messages following its current schema

  #13243 commented on Aug 6, 2025 • 6 new comments

• fix: improve Gemini API key masking in debug logs

  #13272 commented on Aug 6, 2025 • 1 new comment

• [Bug]: OSError: [Errno 24] Too many open files

  #13220 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Run openHands with bedrock

  #10609 commented on Aug 6, 2025 • 0 new comments

• [Bug]: DynamoDB integration error on embedding model

  #10426 commented on Aug 6, 2025 • 0 new comments

• [Bug]: Container fails when redis semantic cache is configured

  #10352 commented on Aug 6, 2025 • 0 new comments

• [Feature]: Supprt for Vertex AI with global region

  #9234 commented on Aug 6, 2025 • 0 new comments

• [Bug]: error while accesing usage section in UI

  #6348 commented on Aug 6, 2025 • 0 new comments

• New Models/Endpoints/Providers

  #4922 commented on Aug 6, 2025 • 0 new comments

• [Feature]: Add support for Llamafile provider

  #3225 commented on Aug 6, 2025 • 0 new comments

• [Feature]: Support `think` parameter for Ollama models

  #11680 commented on Aug 6, 2025 • 0 new comments

• add Lite llm docker proxy (Gemini ver)

  #2574 commented on Aug 5, 2025 • 0 new comments

• 1215同步

  #7243 commented on Aug 5, 2025 • 0 new comments

• Remove files generated by FE builds

  #10067 commented on Aug 6, 2025 • 0 new comments

• Update Model Pricing Information for Groq Llama 4 models

  #10273 commented on Aug 6, 2025 • 0 new comments

• Fix AzureChatCompletion adding stream_options when stream is False

  #10594 commented on Aug 6, 2025 • 0 new comments

• fix(litellm/caching/caching_handler.py): fix kwargs[litellm_params][p…

  #10612 commented on Aug 5, 2025 • 0 new comments

• [Feature: Added the support to add api_url in custom_guardrail]

  #12512 commented on Aug 5, 2025 • 0 new comments

• Fix additional anyOf corner cases for Vertex AI Gemini tool calls - issue #11164

  #12797 commented on Aug 6, 2025 • 0 new comments

• fix: enable metadata parameter for OpenAI API calls

  #12999 commented on Aug 5, 2025 • 0 new comments

• Update Pangea Guardrail to support new AIDR endpoint

  #13160 commented on Aug 6, 2025 • 0 new comments

• [Bug]: Job "update_spend" raised an exception

  #13204 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Http 500 error when trying to call dall-e-3

  #13254 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Total tokens do not count cache create input tokens (LiteLLM Proxy)

  #13207 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Edited `Model Info` fields via UI are not persisted

  #13082 commented on Aug 5, 2025 • 0 new comments

• [Bug]: o1-pro and Cohere models returns "finish_reason": "stop" when using streaming

  #12862 commented on Aug 5, 2025 • 0 new comments

• Intermittent API Connection Errors while sending the async requests to OpenAI (More with the reasoning models)

  #12807 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Copy to Clipboard does not work

  #12474 commented on Aug 5, 2025 • 0 new comments

• [BUG]: Bedrock streaming responses buffer in 1024-byte chunks causing choppy output

  #11747 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Heavy RAM Usage over time

  #12685 commented on Aug 5, 2025 • 0 new comments

• [Feature]: Batch Inference support for AWS Bedrock

  #12681 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Error when using bind_tools() with a fine-tuned gemini-2.0-flash-001 model hosted on GCP

  #12001 commented on Aug 5, 2025 • 0 new comments

• [Feature]: Add support for custom OAuth2-based LLM provider

  #12367 commented on Aug 5, 2025 • 0 new comments

• [Feature]: Support MCP Server instructions in the gateway

  #13119 commented on Aug 5, 2025 • 0 new comments

• [Bug]: Tool-Calling of Anthropic via Bedrock not working: Expected 'id' to be a string

  #13124 commented on Aug 6, 2025 • 0 new comments

• Metadata dict unsupported in `token_counter` since v1.68.0

  #10623 commented on Aug 6, 2025 • 0 new comments

• [Bug]: OTEL authentication header for `arize_phoenix` is missing

  #10621 commented on Aug 6, 2025 • 0 new comments

• [Feature]: Add Cost Tracking for Mistral OCR Models

  #10620 commented on Aug 6, 2025 • 0 new comments

• [Bug]: HF inference provider Cohere `huggingface/cohere/CohereLabs/aya-expanse-32b` calls incorrect route

  #10619 commented on Aug 6, 2025 • 0 new comments

• How to config 3rd part API provider use tokenizer base on model name ?

  #10614 commented on Aug 6, 2025 • 0 new comments

• [Bug]: AttributeError: 'function' object has no attribute 'create' [client]

  #10610 commented on Aug 6, 2025 • 0 new comments