Qwen 3.6 27B: what 76 developers actually say.
76 Hacker News opinions from 58 developers, Nov 2025 – Nov 2026. 63 come from people who say they use it. Here is what they praise, what they complain about, and who switched.
The case for Qwen 3.6 27B, in one paragraph.
Qwen 3.6 27B earns 76% approval across 76 opinions. Developers praise its code quality. In the last 90 days it sits at 56%.
Pick it if you
- Want a capable coding model you can self-host
- Need private inference on your own hardware at low cost
- Do basic editing, scripting, and debugging on a local GPU
Look elsewhere if you
- Need reliable long agentic coding sessions
- Need fast throughput for interactive coding agents
- Need frontier-level coherence on complex or novel work
Where it wins and loses.
Mentions across 76 opinions. Bars are counts, not votes; a tool can be both praised and criticized on the same topic.
View data table
| Topic | Praise | Complaints | Mentions |
|---|---|---|---|
| Code quality | 21 | 4 | 25 |
| Privacy & local use | 15 | 0 | 15 |
| Speed | 8 | 3 | 11 |
How opinion has moved.
Monthly approval, Apr 2026 – Aug 2026. Bars show opinion volume; low-volume months are still included but faded.
View data table
| Month | Approval | Opinions |
|---|---|---|
| Apr 2026 | 67% | 9 |
| May 2026 | 77% | 15 |
| Jun 2026 | 93% | 28 |
| Jul 2026 | 62% | 16 |
| Aug 2026 | 88% | 4 |
What developers say.
Summaries are paraphrased. Each links to the original Hacker News comment.
Praise
4 selectedSelf-hosted Qwen 3.6 27B gives about the same results as Sonnet 4.6 at very low cost.
Qwen 3.6 27B was an easy daily driver, rarely needing larger models except for planning.
Dense Qwen 3.6 27B built a working whiteboard app using only error-log feedback.
Uses Qwen 3.6 27B as their main workhorse on a local GPU.
Complaints
4 selectedThe local Qwen 3.6 27B model was far behind Cursor's default and Sonnet.
Qwen 3.6 27B produced unusably bad results in their testing.
Qwen 3.6 27B often fails to finish their agent-harness optimization task within 100 turns.
Qwen 3.6 27B often can't finish their agent optimization task within 100 turns.
Questions developers ask.
Is Qwen 3.6 27B worth it?
Is Qwen 3.6 27B better than Claude Sonnet?
Can Qwen 3.6 27B run locally?
The data behind this page.
Developers give Qwen 3.6 27B 76% approval across 76 opinions from 58 authors, 63 of them firsthand. Praise centers on code quality (21 positive, 4 negative) and local privacy (15 positive, 0 negative), with 6 positive cost comments and no cost complaints. The trend is down: 90-day approval is 56%, and monthly approval fell from 93% in June to 50% in September and 0% in October on just 2 opinions. Complaints cluster around reliability (5 negative, 4 positive), slow speed, and weak performance in long agent runs.
Every number comes from public Hacker News comments collected through the HN Search API. A language model tagged each comment for sentiment, first-hand use, topic and switches; a human spot-checks samples. No vendor can pay for placement.
Get the monthly report
Which AI coding tools gained or lost developer approval on Hacker News, and why people switched. One email a month.