Ollama
Ollama runs open LLMs locally or in the cloud with CLI and API access.
Tested on LinuxKnown vulnerabilities
Use it
Before you install: known vulnerabilities
pickfit found 12 known high or critical vulnerabilities in the packages a fresh install brings (OSV.dev, 5 Nov 2026). Weigh this before you install it.
npm:seroval 1.3.2seroval: `seroval.fromJSON()` Promise resolver type confusion invokes attacker-controlled methods during deserialization (GHSA-mv8w-475r-vwqw, critical, fixed in 1.5.3)npm:dompurify 3.3.0DOMPurify IN_PLACE: hook that removes a parent or sibling leaves the detached subtree's event handlers armed (incomplete fix for GHSA-p98j-92pf-mc4p) (GHSA-mrv2-3pqr-j5jr, high, fixed in 3.4.17)
See 8 more
npm:seroval 1.3.2— GHSA-p6vx-979v-rg4c (critical), fixed in 1.6.2npm:lodash-es 4.17.21— GHSA-r5fr-rjxr-66jc (high), fixed in 4.18.0npm:seroval 1.3.2— GHSA-3j22-8qj3-26mx (high), fixed in 1.4.1npm:seroval 1.3.2— GHSA-3rxj-6cgf-8cfw (high), fixed in 1.4.1npm:seroval 1.3.2— GHSA-66fc-rw6m-c2q6 (high), fixed in 1.4.1npm:seroval 1.3.2— GHSA-7rv5-mrxv-xcr9 (high), fixed in 1.6.8npm:seroval 1.3.2— GHSA-97r2-qh9m-474h (high), fixed in 1.6.8npm:seroval 1.3.2— GHSA-hj76-42vx-jwp4 (high), fixed in 1.4.1
2 more findings not listed here.
Linux x86_64
Manual install ollama binary to ~/.local without sudo by extracting ollama-linux-amd64.tar.zst
curl -fsSL https://ollama.com/download/ollama-linux-amd64.tar.zst | zstd -d | tar -xf - -C /tmp && mkdir -p ~/.local/bin ~/.local/lib && cp /tmp/bin/ollama ~/.local/bin/ && cp -r /tmp/lib/ollama ~/.local/lib/ 2>&1; ~/.local/bin/ollama --version
Check that it works
~/.local/bin/ollama --version 2>&1; echo "---"; ~/.local/bin/ollama --help 2>&1The output should contain “Large language model runner”.
First thing to try
LD_LIBRARY_PATH=$HOME/.local/lib/ollama:$LD_LIBRARY_PATH $HOME/.local/bin/ollama servecurl -s http://localhost:11434/api/tags
echo "---VERSION---"
curl -s http://localhost:11434/api/versionWhat it does
- Run and chat with open modelsTested
- Run and manage models via REST APITested
- Connect models to existing AI agentsNot shown working
- Turn Ollama into personal AI assistantNot shown working
How far pickfit got
Linux x86_64
- Installs
- Starts
- 2 of 4 features work
- Works offline
Good for
- chat with a downloaded open model
- connect local models to coding agents
- run a personal AI assistant over messaging
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
It produces model output
- Replayed from scratch
- Works offline
- Checked on 4 Nov 2026
- Version 0.40.14
Differences the test machine hit
- README: curl -fsSL https://ollama.com/install.sh | sh installs Ollama
In the test: Install script requires sudo to write to /usr/local; in this sandbox without root manually extracted tarball with curl | zstd -d | tar -xf - and copied bin/ollama to ~/.local/bin with LD_LIBRARY_PATH to ~/.local/lib/ollama