Ollama
Free, private way to run open coding models like Qwen3.8 27B on your own computer; skip it if you need frontier quality.
Should you use it?
Pick it if
People who want coding help that never leaves their machine, and developers who want a local model API.
Best for: Developers who want a local model API, and private coding help (Run AI on your computer Top 20)
Skip it if
You want the best coding quality; free Claude (Sonnet 5.5) ranks far higher on WebDev, as does GPT-6.1 Sol on ChatGPT Plus.
You give up: Slower than mlx-lm on a Mac in one test; its bigger models are cloud-only
The short version
- What this is
- A free, open-source tool that downloads open AI models and runs them on your own computer, with optional paid cloud models.
- Why it matters
- It keeps your code and chats on your own machine at no cost, which matters for private or offline work.
- Free and unlimited for local use; MIT licensed s4 s7.
- Qwen3.8 27B, an 18GB download in its library, ranks 28th on arena.ai WebDev s3 s2.
- Easiest setup, and as fast as plain llama.cpp in one Mac test, though mlx-lm was faster there s1.
Good at
Weak at
- On a Mac it was slower than mlx-lm (77.7 vs 107.5 tokens a second), and its large default context used 15 GB of memory for a 6.6 GB model, in one independent test s1.
- Local models sized for a normal computer sit well below the leaders: Qwen3.8 27B is rank 80 of 409 on arena.ai Text Coding s8.
- Bigger models such as GLM-5.3 and DeepSeek V4.1 Flash are cloud-only in its library s2.
Our take
Ollama is the simplest free way to run an open model on your own computer, and nothing you type leaves the machine s5. The score is held back by the models, not the tool: Qwen3.8 27B, an 18GB coding model in its library s2, ranks 28th of 142 on WebDev s3, while free cloud options such as Claude's Sonnet 5.5 rank near the top.
Speed is close to the engine underneath. In TerminalBytes' Mac test it matched plain llama.cpp, while Apple's mlx-lm was about 40% faster; Ollama's large default context also used more memory than the model file needs s1. For most people the easy setup is worth it. The 18GB figure is the download size; Ollama's FAQ gives no minimum hardware, so check the model page against your machine first s5.
Score
A score from 0 to 100 per job, against everything available today. Overall 80 across 1 job, the average of the job scores.
New to this? What the scores and labels mean
- 0 to 100
- Our score for one job, like writing or coding, compared with everything else out today. 100 is the best available today. For AI models we build it from public data, such as Arena's blind votes and Epoch AI's independent tests; apps and tools are scored on our own checklist. The same tool can score differently for different jobs, and where two likely ranges overlap, we call it a tie.
- Our verdict
- Use it: a top choice for most people. Worth trying: good, and worth a go. Depends: good for some people or one kind of job. Wait: too early to tell, or not out for everyone yet. Skip: something else does the job better.
- First look
- Just released. We say what it is and whether to bother, and give it a score once public results exist.
- No score yet
- None of the public sources we use covers it yet, or the only evidence is from a source whose numbers we can't use, so we describe it in words instead.
- Best buy
- Nothing cheaper scores as well for that job.
- Best free
- The best score for that job that costs nothing, through a free plan or by running on your own computer.
- Pay as you go
- No monthly plan: you pay for each use, usually through a developer account.
- Checked
- The day we last confirmed the score and the price against the sources. Prices change often, so check the date.
- AI model
- The engine underneath, like GPT or Claude. You use it through an app or, if you code, directly.
- Open source
- Free to download and run on your own computer, so nothing leaves your machine.
Our checklist score, going by an independent test, a public leaderboard and the maker’s pages.
Our checklist
2 sources
Sources and notes
Independent testPublic leaderboardMaker’s pagesChecklist 8/10: does the job 1 (TerminalBytes, 20 Sep 2026, Ollama 0.32.13 on a Mac Studio M3 Ultra: 77.7 tokens a second, level with plain llama.cpp at 77.8, but behind mlx-lm at 107.5, and its default 262,144-token context raised memory use to 15 GB for a 6.6 GB model, s1; speed only); underlying quality 1 (its library carries qwen3.8:27b, an 18GB download, s2, which is arena.ai WebDev rank 28 of 142, qwen3.8-27b, s3); value 2 (local use free and unlimited, s4); privacy 2 (Ollama does not see prompts run locally, s5); limits 2 (v0.40.2 released 8 Oct 2026, s6).
Alternatives
Better or cheaper for coding, from our scores, then the ones we'd also look at. The same kind of product comes first.
- Claude Opus 5.5BetterText model100$20 a monthScores 100 in coding, against 80 here; $20 a month. Side by side
- llama.cppLibraryNot reviewed yetThe engine underneath; the same speed in TerminalBytes' Mac test, with more control.
- LM Studio (with Bionic agent)App80FreeA desktop app with a model browser and the Bionic agent, if you prefer not to use the terminal. Side by side
- ClaudeApp80FreeMuch stronger coding model for $0, if sending code to a cloud service is fine. Side by side
Value for money
Where it sits in coding: higher is a better score, further left is cheaper. Green dots are the Best buys.
Evidence
Everything behind the call and the scores, folded so the page stays short. Spotted a wrong price, score or link? Report a mistake.
How we scored this
The coding score comes from our own checklist of the product.
Independent testPublic leaderboardMaker’s pages
Reviews are drafted with help from AI models, including Anthropic's Claude. Model scores come only from outside leaderboards and independent evaluations, by a written rule.
No one paid for this review or gave us free access, and we earn nothing if you sign up. How we score
Sources 8
- s1TerminalBytes: Ollama vs llama.cpp vs MLX on Mac (2026), 20 Sep 2026
- s2Ollama model library (newest) and qwen3.8 model page
- s3arena.ai Code Arena: WebDev leaderboard
- s4Ollama pricing
- s5Ollama FAQ
- s6Ollama releases on GitHub
- s7Ollama repository on GitHub
- s8arena.ai Text: Coding leaderboard
Something here out of date or wrong? Report a mistake and we will check it against the source.
History 3 changes
Every change to a rating, the price or our call, with the reason. Scores are against what exists on the day, so they drift down as better things ship. Older entries come from before we moved to scores.
- Evidence swapped: the InventiveHQ test was first published in June, outside our 90-day window; TerminalBytes (20 Sep 2026, Ollama 0.32.13) replaces it. Rating unchanged.
- New: Coding ratedReviewed: coding (v0.40.2).
- Audit: evidence now says the InventiveHQ measurement is speed only, which is why 'does the job' is 1; removed the unsourced claim that Qwen3.8 27B is the best coding model that fits most computers. Rating unchanged.
Use it well 4 guides on this site
Pages on this site that help with the jobs above.