Owner

ollama

3 indexed repositories · View on GitHub

ollama
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

Go · 181247 stars
ollama-python
Ollama Python library

Python · 10537 stars
ollama-vscode
Official Ollama Extension for VS Code

TypeScript · 52 stars

Beginner friendly Everything

Open All

MLX: mlx-community Gemma 4 MoE imports but fails to load: layer 0: missing MoE expert weights (experts.switch_glu layout) Open

Difficulty 3/5 1-2 days Newbie friendliness 78/100

ollama/ollama#18540 · Opened Sep 19, 2026
Upgrading to Ollama 0.34.2 breaks the extension Open

Difficulty 4/5 3-5 days Newbie friendliness 45/100

ollama/ollama-vscode#70 · Opened Sep 19, 2026
OpenAI-compatible endpoint silently ignores `reasoning_content` on assistant messages, dropping replayed reasoning for DeepSeek models Open

Difficulty 3/5 1-2 days Newbie friendliness 58/100

ollama/ollama#18534 · Opened Sep 19, 2026
Vulkan on Intel Iris Xe iGPU: qwen2.5:14b fails with ErrorOutOfDeviceMemory (KV cache alloc) on 0.33.3 and 0.34.2; works on 0.32.6 Open

Difficulty 4/5 3-5 days Newbie friendliness 48/100

ollama/ollama#18531 · Opened Sep 18, 2026
qwen3-coder: tool call silently lost when the model puts reasoning before it (model omits <tool_call> opener; parser only keys on the literal tag) Open

Difficulty 3/5 1-2 days Newbie friendliness 75/100

ollama/ollama#18530 · Opened Sep 18, 2026
qwen3.5-family hybrid GDN models on ROCm: cross-request state leak in the bundled llama-server (0.32.14 and 0.34.1) — earlier prompts' text appears in later responses; tracking ggml-org/llama.cpp#29092 Open

bug

Difficulty 5/5 Over a week Newbie friendliness 38/100

ollama/ollama#18528 · 6 comments · Opened Sep 18, 2026
gpt-oss:20b (MXFP4) deterministic llama-server abort in CUDA ADD_ID on a short two-message /api/chat (0.32.13 and 0.33.3) Open

bug nvidia

Difficulty 4/5 3-5 days Newbie friendliness 45/100

ollama/ollama#18522 · 1 comment · Opened Sep 18, 2026
Support Prism ternary GGUFs (PQ2_0 type 142 / PTQ1_0 type 143): import fails with unsupported tensor size overflows Open

model

Difficulty 4/5 Over a week Newbie friendliness 35/100

ollama/ollama#18521 · 1 reaction · Opened Sep 18, 2026
Proposal: Add `--nodraft` flag to `ollama run` to disable speculative decoding Open

feature request

Difficulty 3/5 1-2 days Newbie friendliness 68/100

ollama/ollama#18517 · 1 comment · Opened Sep 17, 2026
Add MLX support for Bonsai's low-bit (1-bit/2-bit) quantized weights to the new 0.19 MLX backend. Open

model

Difficulty 4/5 3-5 days Newbie friendliness 48/100

ollama/ollama#18515 · 3 reactions · Opened Sep 17, 2026
ollama cloud settings panel login with anonaddy.me email alias: Access blocked, please contact support. Open

bug

Difficulty 5/5 Over a week Newbie friendliness 20/100

ollama/ollama#18513 · 1 comment · Opened Sep 17, 2026
Ollama refusing tool role, which always worked fine in llama.cpp with qwen. Open

bug needs more info

Difficulty 4/5 3-5 days Newbie friendliness 38/100

ollama/ollama#18509 · 14 comments · Opened Sep 17, 2026
glm-5.3-flash intermittently emits malformed string-encoded tool calls via Ollama Cloud Open

cloud

Difficulty 4/5 3-5 days Newbie friendliness 35/100

ollama/ollama#18506 · Opened Sep 17, 2026
[BUG] MLX nvfp4: admitted request stalls in prefill at processed=total-1 with zero tokens for minutes under sustained single-slot load; only runner SIGTERM recovers Open

bug mlx

Difficulty 4/5 3-5 days Newbie friendliness 48/100

ollama/ollama#18505 · 3 comments · Opened Sep 17, 2026
linux: fix Vulkan inference support on ARM64 (PR implemented) Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 15/100

ollama/ollama#18502 · Opened Sep 17, 2026
qwen3-vl:8b-instruct 0xc0000005 on Vulkan AMD RX 6750 XT after multi-model load (Windows 0.34.1) Open

vulkan windows

Difficulty 4/5 3-5 days Newbie friendliness 35/100

ollama/ollama#18494 · Opened Sep 16, 2026
Connect to Claude Desktop forks into a new anonymous account instead of the signed-in session, and Disconnect intermittently fails to restore it Open

Difficulty 4/5 3-5 days Newbie friendliness 38/100

ollama/ollama#18491 · 1 comment · Opened Sep 16, 2026
Feature request: optional external resource lock for shared GPU coordination Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 35/100

ollama/ollama#18487 · 2 comments · Opened Sep 16, 2026
Model not showing up in picker even after successful diagnostics Open

ollama/ollama-vscode#69 · 1 comment · 1 assignee · Opened Sep 16, 2026
Vulkan iGPU (Intel Iris Xe) not detected at startup - total_vram 0 B sticky until restart Open

vulkan

Difficulty 4/5 3-5 days Newbie friendliness 48/100

ollama/ollama#18482 · Opened Sep 16, 2026
typical_p is no longer supported (status code: 400) from python API Open

bug

ollama/ollama#18480 · 1 assignee · Opened Sep 15, 2026
[Bug]: Vulkan iGPU runner wedges after a cancelled large prefill — later requests hang silently until backend restart Open

vulkan

Difficulty 4/5 3-5 days Newbie friendliness 35/100

ollama/ollama#18477 · 1 comment · Opened Sep 15, 2026
Extremely slow response times when using Claude integration Open

needs more info

Difficulty 4/5 3-5 days Newbie friendliness 45/100

ollama/ollama#18474 · 1 comment · Opened Sep 15, 2026
gemma4 renderer drops tool parameters named `description`, `type`, `properties`, `required` or `nullable` Open

Difficulty 3/5 1-2 days Newbie friendliness 78/100

ollama/ollama#18468 · 1 comment · Opened Sep 15, 2026
`ollama launch claude` starts 1M-context cloud models with a 200K window, and labels a 262K model as 1M Open

Difficulty 3/5 1-2 days Newbie friendliness 65/100

ollama/ollama#18463 · Opened Sep 15, 2026
Intermittent "model xxx not found" errors Open

bug

Difficulty 2/5 1-3 hours Newbie friendliness 20/100

ollama/ollama#18447 · Opened Sep 14, 2026
Add support for Qualcomm IQ-9075 NPU/GPU Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 25/100

ollama/ollama#18445 · 2 comments · Opened Sep 14, 2026
gemma4:26b concurrent decode loses EOS (length at num_predict); qwen3.8-27b IDENTICAL dual on same GB10/Ollama harness is 3/3 clean Open

bug

Difficulty 5/5 Over a week Newbie friendliness 35/100

ollama/ollama#18442 · 1 comment · Opened Sep 14, 2026
MLX: structured output with thinking enabled prefixes JSON content with a stray "." Open

mlx

Difficulty 4/5 3-5 days Newbie friendliness 65/100

ollama/ollama#18441 · 1 comment · Opened Sep 14, 2026
convert_function_to_tool exposes *args and **kwargs as required string parameters Open

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

ollama/ollama-python#735 · Opened Sep 14, 2026
A single image passed as a bare string/bytes is iterated character-by-character Open

Difficulty 2/5 1-3 hours Newbie friendliness 72/100

ollama/ollama-python#734 · Opened Sep 14, 2026
ChatRequest.model_dump() crashes when a tool has no parameters or no function Open

Difficulty 2/5 1-3 hours Newbie friendliness 75/100

ollama/ollama-python#733 · Opened Sep 14, 2026
Support ROCm 10 for Windows Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 25/100

ollama/ollama#18435 · Opened Sep 14, 2026
Anthropic-compat: system-role messages inside `messages` are hoisted into the system block, defeating the prefix cache (Claude Code) Open

Difficulty 3/5 1-2 days Newbie friendliness 74/100

ollama/ollama#18431 · 3 comments · Opened Sep 13, 2026
qwen3-coder: extra tool schema keys render in random order, so identical requests miss the prompt cache Open

Difficulty 2/5 1-3 hours Newbie friendliness 78/100

ollama/ollama#18430 · Opened Sep 13, 2026
Hi Ollama Open

Difficulty 5/5 Over a week Newbie friendliness 20/100

ollama/ollama#18429 · 1 comment · Opened Sep 13, 2026
Request to add Gnani Evon-v3.3 Open

model

Difficulty 4/5 3-5 days Newbie friendliness 35/100

ollama/ollama#18427 · 1 comment · Opened Sep 13, 2026
kimi-k3:cloud crashes (HTTP 500) on image content in tool-role messages — regression from kimi-k2.6 + glm-5.3-flash both work Open

cloud

Difficulty 4/5 3-5 days Newbie friendliness 45/100

ollama/ollama#18426 · Opened Sep 13, 2026
Docker SBX Support for ollama integration agents Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 45/100

ollama/ollama#18425 · Opened Sep 13, 2026
Qwen3-Coder tool parser changes number arguments outside int64 range Open

Difficulty 2/5 1-3 hours Newbie friendliness 78/100

ollama/ollama#18421 · Opened Sep 13, 2026
/api/codex/v1/responses silently returns empty completion for previous_response_id tool follow-up Open

ollama/ollama#18419 · 2 comments · 1 assignee · Opened Sep 13, 2026
[BUG]: Model sees rotated images Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 64/100

ollama/ollama#18418 · Opened Sep 13, 2026
`ollama create --quantize` from safetensors leaves the unquantized F16 blob in `blobs/` (unreferenced, never removed) Open

Difficulty 4/5 3-5 days Newbie friendliness 58/100

ollama/ollama#18416 · Opened Sep 13, 2026
Windows 10: brief visible PowerShell window when opening Ollama desktop 0.34.0 Open

Difficulty 4/5 3-5 days Newbie friendliness 48/100

ollama/ollama#18415 · 1 reaction · Opened Sep 12, 2026
Some models have undocumented version requirements Open

bug ollama.com

Difficulty 4/5 3-5 days Newbie friendliness 45/100

ollama/ollama#18414 · 2 reactions · Opened Sep 12, 2026
Linux hybrid graphics (Intel Raptor Lake-S iGPU + NVIDIA RTX 4080): llama-server crashes with SIGABRT during backend/device loading Open

Difficulty 4/5 3-5 days Newbie friendliness 45/100

ollama/ollama#18412 · 1 comment · Opened Sep 12, 2026
Responses: web_search path emits function_call before reasoning completion, breaking Codex tool replay Open

Difficulty 2/5 1-3 hours Newbie friendliness 90/100

ollama/ollama#18411 · Opened Sep 12, 2026
Jetson Orin Nano 8GB: Gemma 4 E4B OOM with --load-mode dio, while identical configuration succeeds without DIO Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 48/100

ollama/ollama#18396 · 9 comments · Opened Sep 11, 2026
Include the served manifest digest in local /api/chat responses (tested v0.34.0 patch) Open

Difficulty 3/5 1-2 days Newbie friendliness 74/100

ollama/ollama#18394 · 1 comment · Opened Sep 11, 2026
Sustained `/api/embed` load exhausts loopback ports on Windows: llama-server HTTP client has keep-alive disabled Open

Difficulty 3/5 1-2 days Newbie friendliness 72/100

ollama/ollama#18392 · 1 comment · Opened Sep 11, 2026

Showing the newest 100

This page lists what was indexed most recently. Advanced filter has the whole inventory, narrowed by language, difficulty and how long a task takes.

Open advanced filter

ollama

Showing the newest 100

Get new issues in your inbox