Problemsby GlyphOllama upgrades break working setups
21 posts from 17 people17d activeFading, +0 this week
Posts per day
The posts behind it
cloud: /api/usage no longer returns session/weekly usage percentages, only request counts
## Summary `GET https://ollama.com/api/usage` (Bearer API key) no longer returns the session/weekly usage percentages. It now returns only request-count buckets. The percentages still exist in the web UI at https://ollama.com/settings, so this looks like a regression rather than...
Can an older library manifest still be pulled by digest after its tag is updated?
I accidentally re-pulled qwen3.5:9b and its tag now points to a newer manifest. Today, ollama show qwen3.5:9b --modelfile / /api/tags reports digest 56671c2ab9385f9cfcb404638e32cd62d88e3501d44822208363c010179a3c90. Before the pull, my local digest started with 6488c96fa5fa (I...
[Bug]: v1.10.0 AI requests fail with "fetch failed" (undici 8 Agent passed to Node 24's built-in fetch)
### Are you willing to submit a PR? No, but I can help test ### What happened? After upgrading to v1.10.0, every AI request to my Ollama service fails right away with `502 {"error":"Failed to reach the AI service: fetch failed","code":"UPSTREAM_ERROR"}`. It answers in about 0.2...
issue: Backspace during IME composition force-commits the text; in Firefox it also copies the composing string to the start of the chat input
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
ROCm on Radeon 890M (gfx1150, iGPU): garbage output and "Memory access fault by GPU" on 0.40.6 and 0.40.7; 0.33.3 works
...7.2.4-1, `rocblas` 7.2.4-3, `rocm-core` 7.2.4-1), same in the working and failing setups - Ollama: Arch `ollama` / `ollama-rocm` 0.33.3-1 (works), 0.40.6-1 and 0.40.7-1 (fail) - Service environment: `OLLAMA_IGPU_ENABLE=1`, `OLLAMA_FLASH_ATTENTION=1`,...
[Bug]: Ollama legacy URLs drop proxy path prefixes
### What happened? `OllamaEmbeddingFunction` accepts the legacy `/api/embeddings` endpoint for backward compatibility, but its conversion to a base URL keeps only the scheme and authority. A reverse-proxy path prefix is lost. For example,...
olmo3-think: content starts with the newlines after </think> because the tag and the whitespace arrive in different chunks
### What is the issue? With the `olmo3-think` parser (OLMo 3 Think and OLMo 3.1 Think), every response's content begins with the whitespace the model emits after ` `, typically `\n\n`. The thinking field is clean; the content is not. The parser trims that whitespace only when it...
issue: The rebuilt search dialog no longer understands any of its five filters, and archived chats cannot be found
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
issue: searching for a tag whose chats are all archived loses the tag's name
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
issue: a collaborator renaming a shared knowledge base turns off its Public sharing
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
Windows: tray app starts "ollama serve" about once per second, forever, when a server already listens on 11434
...different trigger: #12206 shows the same `ollama exited` loop signature when the port is blocked by the system (bind error "forbidden by its access permissions"). #14761 describes the app respawning `ollama serve`. ### Relevant log output `app.log`: `server.log` (repeated once...
issue: a calendar event moved to a new time never reminds me again
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
issue: [regression] after a socket drop, a tab that signed in without reloading reconnects in a loop (server disconnects, client reconnects at once)
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
Go API: tool-call arguments silently lose numeric precision
Proposed fix and regression tests: https://github.com/ollama/ollama/pull/18941 ## What is the issue? Numeric tool-call arguments lose precision when decoded by the Go API client. This occurs in both normal and streaming `Client.Chat` responses. Reusing the returned assistant...
MLX runner panic with qwen3.6:35b-mlx in Ollama 0.40.x — regression from 0.35.0
...failed: panic: cache: offset 5335 exceeds 5334 stored keys` LiteLLM reports it as: `Unable to parse ollama chunk` The failure occurs while generating the first report section, after the application has successfully loaded and indexed the paper and assembled the evidence. The...
issue: Creating a model in Manage Ollama shows no progress when Ollama downloads its base model
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
[BUG] Reasoning plan misses bare READY verdicts, loops forever (refs #6204)
## Description All three planning paths in `lib/crewai/src/crewai/utilities/reasoning_handler.py` (`_call_with_function` x2, `_parse_planning_response`) detect readiness with only the literal `"READY: I am ready to execute the task."`. Models that emit a bare `READY` verdict...
Misc. bug: High CPU use with model running on GPU. Regression introducted in b10423)
### Name and Version I am building llama.cpp from git repo. Recent versions of llama-server use excessive CPU when running models entirely on GPU. I have done git bisect and this is the first bad commit: Prior to this running models like Gemma 4 26b uses very low CPU, but after...
Ollama 0.35.0 regression: model pulls bypass HTTPS_PROXY for Cloudflare R2 downloads
### What is the issue? ### What is the issue? After upgrading the Ollama server from 0.34.4 to 0.35.0, model pulls no longer work in an environment where outbound HTTPS access must go through an HTTP forward proxy. The manifest request succeeds, but when Ollama attempts to...
issue: the sidebar stops loading older chats after a new chat or an import, until the page is refreshed
### Before Submitting - [x] I searched open and closed issues and discussions for an existing report. - [x] I reproduced this bug on the latest release AND on the current `dev` branch, right before submitting this report. I did not just check an old version or rely on a check...
Vulkan backend fails with 'Not enough memory for command submission' loading a 66 GB model on AMD iGPU (regression since v0.32.12)
.../ v0.33.3) ### Notes - Same hardware, same config, same env vars in both working and failing versions - only the ollama binary/libs changed (installed the same way, replacing /usr/local/bin/ollama and /usr/local/lib/ollama). - The 66 GB model only fits because of the CPU/GPU...
Companies and products named
About this problem
Evidence
21 posts from 17 people in 6 places, about 2 a week over 60 days. Mostly on ollama/ollama, open-webui/open-webui, ggml-org/llama.cpp. Tools named alongside: Open WebUI, GitHub, Docker, llama.cpp.
Frustration Frustration 0 of 3· Seen on GitHub issues
How it was grouped
Posts that state a pain and match the "breaking-changes" rule about Ollama. First post Sep 6, 2026, latest Nov 4, 2026. Corroborated: 3 or more posts from 2 or more people or places. Method
The same problem elsewhere
GitHub upgrades break working setups
89 posts from 77 people on GitHub issues, Hacker News, Stack Overflow.
Upgrades and new releases break working setups
77 posts from 63 people on GitHub issues, Reddit, App Store reviews.
Claude Code upgrades break working setups
68 posts from 61 people on GitHub issues, Hacker News, Reddit.
Codex upgrades break working setups
67 posts from 60 people on GitHub issues, Reddit.
DuckDB upgrades break working setups
45 posts from 35 people on GitHub issues.
Claude upgrades break working setups
35 posts from 34 people on GitHub issues, Reddit, Hacker News.
ClickHouse upgrades break working setups
31 posts from 20 people on GitHub issues.
Anthropic upgrades break working setups
27 posts from 22 people on GitHub issues.
PyTorch upgrades break working setups
26 posts from 21 people on GitHub issues, Reddit.
ChatGPT upgrades break working setups
26 posts from 26 people on GitHub issues, Community forums, App Store reviews.
OpenAI upgrades break working setups
25 posts from 25 people on GitHub issues, Reddit, Community forums.
Open WebUI upgrades break working setups
25 posts from 2 people on GitHub issues.
Other Ollama problems
Running Ollama locally hits memory and speed limits
47 posts from 45 people on GitHub issues, Reddit.
Ollama users keep filing bugs
Top post: "Ollama does not work even after a fresh install", one of 78 posts from 48 people on GitHub issues, Reddit, Community forums.
Ollama error messages are cryptic
13 posts from 7 people on GitHub issues, Reddit.
Ollama runs out of context on real work
7 posts from 7 people on GitHub issues, Reddit.
Trouble with Ollama empty
Top post: "Ollama photo/label estimation returns empty JSON with reasoning models (gemma4, qwen3.6), request should send", one of 7 posts from 6 people on GitHub issues.
Serving models on Ollama is harder than it should be
6 posts from 6 people on GitHub issues, Reddit.
Trouble with Ollama mode
Top post: "auto-playback should read responses sentence by sentence while streaming, like call mode", one of 6 posts from 4 people on GitHub issues.
Trouble with Ollama pull
Top post: "Failed to pull `embeddinggemma-2:740m` on linux", one of 6 posts from 6 people on GitHub issues.
Trouble with Ollama local
Top post: "issue: MinerU local engine gets 401 from a MinerU 4.x server started with --api-key, even with the API key set", one of 4 posts from 4 people on GitHub issues.
Ollama fails to return valid structured output
4 posts from 3 people on GitHub issues.
Trouble with Ollama text
Top post: "Ollama Text Classifier modles slower under version 0.40.4 ?", one of 4 posts from 3 people on GitHub issues.
Trouble with Ollama instead
Top post: "issue(a11y): pressing Enter on a notification's Dismiss button opens the chat instead of dismissing it", one of 4 posts from 3 people on GitHub issues.
History