Trouble with OpenAI responses
19 posts from 17 people16d activeFading, +0 this week
Posts per day
The posts behind it
[Bug]: glm47 tool parser silently drops tool calls with unknown function names, returning empty responses that trap clients in retry loops
### Your current environment The output of python collect_env.py Note: reproduced on a DGX Spark container image that carries a vLLM dev build (`0.1.dev21830+g8244d1c44.d20261102`); the parser modules involved (`vllm/parser/glm47_moe.py`, `vllm/parser/engine/parser_engine.py`)...
Project Sources fail to save new responses in large Projects
[BUG] OpenAI Responses requests crash on audio inputs advertised by provider constraints
## Description `OPENAI_RESPONSES_CONSTRAINTS` advertises audio support, so an audio file attached to a task is auto-injected into the OpenAI Responses request instead of being handled by the file tools. The Responses formatter, however, only knows how to emit image and PDF...
langchain-openai: Responses API streaming never emits a completed event for function_call
### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...
langchain-openai: Responses API streaming never emits a completed event for function_call
### Package `langchain-openai` ### Description When streaming through the Responses API, `langchain-openai` handles `response.output_item.added` for a `function_call` and every `response.function_call_arguments.delta`, but it does **not** produce a generation chunk for...
langchain-openai: Responses API streaming never emits a completed event for function_call
### Package `langchain-openai` ### Description When streaming through the Responses API, `langchain-openai` handles `response.output_item.added` for a `function_call` and every `response.function_call_arguments.delta`, but it does **not** produce a generation chunk for...
Responses API silently truncates oversized input with truncation disabled (0.40.13 and 0.40.14, Bonsai 27B)
### What is the issue? Observed on Windows with stock Ollama 0.40.13 and independently reproduced with the official portable 0.40.14 release (API version confirmed; archive SHA256 verified). Model: bonsai-2:27b-q2_0, digest...
[Bug]: responses() sends temperature to OpenAI o-series models even with drop_params=True
### Description `litellm.responses(model="openai/o3-mini", temperature=0.5, drop_params=True)` still sends `"temperature": 0.5` to OpenAI. O-series models only accept the default temperature, so the real API rejects the request. The same arguments through `litellm.completion()`...
[Bug]: Playground never shows reasoning on /v1/responses (listens for the removed `response.reasoning.delta` event)
### Description When the Playground's endpoint type is `/v1/responses`, the model's reasoning never shows up. The "Show reasoning" block doesn't appear, even though the response clearly contains reasoning: the metrics row under the answer shows something like `Reasoning: 187`....
[Bug]: completion_cost() raises ModelNotMappedError for groq/openai/gpt-oss-* responses, even when model="groq/..." is passed
### Description `litellm.completion_cost(completion_response=resp)` cannot price a response from `groq/openai/gpt-oss-120b`. LiteLLM itself already priced the call correctly during the request (`resp._hidden_params["response_cost"] == 2.295e-05` for 77 prompt + 19 completion...
[Bug]: Responses API rejects assistant input items without "type" (output_text content)
### Your current environment Docker image: vllm/vllm-openai:v0.31.0 Command: vllm serve Qwen/Qwen3.8-27B --served-model-name hf-qwen38-27b --reasoning-parser qwen3 --tool-call-parser qwen3_coder --enable-auto-tool-choice Also reproduced with GLM-5.3-Flash-W4A16-MTP. ### 🐛...
/v1/responses streaming: created_at changes between response.created and response.completed
### What is the issue? When `/v1/responses` streams, the `response` object inside `response.created`, `response.in_progress` and `response.completed` carries a different `created_at` whenever generation crosses a second boundary. Each event reads the clock when it builds the...
OpenAI-compatible responses use `chatcmpl-<rand.Intn(999)>` as id: 1000 possible values, collisions break clients that key on it
### What happened `middleware/openai.go` sets the response id with `fmt.Sprintf("chatcmpl-%d", rand.Intn(999))` (same for completions/embeddings). Clients that treat the OpenAI `id` as unique (as OpenAI's are) break: LiteLLM's proxy keys its spend-log rows on it and silently...
[FEATURE] Execute hosted-discovery Responses function calls through the authorized tool catalog
## Python PR scope Implement the execution milestone after deferred request building: handle function calls discovered by OpenAI hosted tool search using the caller's existing authorized CrewAI tool catalog. **Difficulty:** hard. **Allocation index:** #8118. **Project:** #8008....
OpenAI-Compatible Endpoint Returns 500 on Responses API, Should Fall Back to Chat Completions (n8n agents)
...N8N_OPENAI_FALLBACK_ENABLED=false - Credential setting: "Disable Responses API fallback" Workaround Until this is fixed, users can manually add apiStyle: 'chat' to the credential JSON in the database: UPDATE credential SET data = jsonb_set(data, '{apiStyle}', '"chat"') }
[Bug]: Azure SDK Responses API calls (/openai/responses) bypass the router and go to api.openai.com via OpenAI pass-through
...and skips `model_list` routing, load balancing, fallbacks and per-model config. ## Workarounds considered - Client-side: call ` /openai/v1/responses` or ` /v1/responses` with a plain `OpenAI` client instead of `AzureOpenAI`. Works, but every Azure-SDK user has to know to do it....
langchain-openai: Responses API drops Azure content_filters from response_metadata
### Checked other resources - [x] I searched existing issues and found nothing on Azure `content_filters` with the Responses API. - [x] The behavior is on `master` (884d2d6) and in langchain-openai 1.6.7. ### Package langchain-openai ### Description Azure OpenAI adds a top-level...
[langchain-openai] ChatOpenAI forwards unsupported sampling parameters to gpt-6-astra via Responses API
### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...
[FR] TypeScript @mlflow/openai: keep tracing spans open until streaming responses are consumed
> [!WARNING] > Before submitting a PR, please make sure that: > - A maintainer has triaged this issue and applied the `ready` label > - This issue has no assignee > - No duplicate PR exists > > PRs not meeting these requirements may be automatically closed. ### Motivation...
Companies and products named
About this problem
Evidence
19 posts from 17 people in 8 places, about 2 a week over 80 days. Mostly on langchain-ai/langchain, BerriAI/litellm, ollama/ollama. Tools named alongside: LangChain, Ollama, GitHub, Anthropic.
Frustration Frustration 0 of 3· Seen on GitHub issues, Community forums
Summary, model-written
Azure OpenAI and LangChain users encounter fallback failures, missing content filter support, and routing issues with the Responses API.
How it was grouped
Posts about OpenAI with no known issue, sharing the word "responses". First post Aug 27, 2026, latest Nov 14, 2026. Corroborated: 3 or more posts from 2 or more people or places. Method
Other OpenAI problems
OpenAI users keep asking for help
Top post: "Multi-file uploads failing and ZIP archives not parsing / sandbox timeout (100% Limits Remaining)", one of 49 posts from 46 people on Community forums, GitHub issues, Reddit.
OpenAI upgrades break working setups
32 posts from 32 people on GitHub issues, Reddit, Community forums.
Trouble with OpenAI chat
Top post: "Chat attachments inside Projects intermittently fail to materialize into /mnt/data, while File Library and", one of 16 posts from 16 people on GitHub issues, Community forums, Reddit.
Running OpenAI locally hits memory and speed limits
14 posts from 14 people on GitHub issues.
OpenAI fails to return valid structured output
12 posts from 12 people on GitHub issues, Hacker News.
Trouble with OpenAI image
Top post: "Image generation is broken for serious work", one of 9 posts from 9 people on GitHub issues, Community forums, Reddit.
OpenAI runs out of context on real work
9 posts from 9 people on GitHub issues, Community forums.
OpenAI rate limits and usage caps block work
7 posts from 7 people on GitHub issues, Community forums.
Trouble with OpenAI calls
Top post: "Unable to connect to dot calls across mobile, web and desktop", one of 7 posts from 6 people on GitHub issues, Community forums.
OpenAI token costs are hard to predict
5 posts from 4 people on GitHub issues, Hacker News, Reddit.
OpenAI error messages are cryptic
5 posts from 5 people on GitHub issues.
Trouble with OpenAI custom
Top post: "docs: Community guide idea, HAL SUPREME as a custom OpenAI-compatible connection", one of 5 posts from 5 people on Community forums, GitHub issues, Reddit.
History