DeepSeek V4 Pro: what 121 developers actually say.

121 Hacker News opinions from 85 developers, Nov 2025 – Nov 2026. 104 come from people who say they use it. Here is what they praise, what they complain about, and who switched.

121 opinions85 developersNov 2025 – Nov 2026Methodology
Approval
75%
±6.5 at 90% confidence
Last 90 days
67%
↓ 8 pts vs all-time
Rank
#10
of 45 models · CI-adjusted
Opinions
121
104 first-hand
Developers
85
1.4 opinions each
The short version

The case for DeepSeek V4 Pro, in one paragraph.

DeepSeek V4 Pro earns 75% approval across 121 opinions. Developers praise its cost & pricing. In the last 90 days it sits at 67%.

Pick it if you

  • Want low cost for heavy daily use
  • Are building hobby or side projects on a budget
  • Want generous usage limits

Look elsewhere if you

  • Need reliable multi-step tool use
  • Cannot absorb rework time
  • Need top-tier code quality

Where it wins and loses.

Mentions across 121 opinions. Bars are counts, not votes; a tool can be both praised and criticized on the same topic.

TopicComplaints ← → PraiseMentions
Cost & pricing 58 mentions
355
Code quality 41 mentions
1031
Reliability 11 mentions
92
PraiseComplaint
3 topics shown · sorted by net sentiment · source: 121 opinions
View data table
TopicPraiseComplaintsMentions
Cost & pricing55358
Code quality311041
Reliability2911

How opinion has moved.

Monthly approval, Apr 2026 – Sep 2026. Bars show opinion volume; low-volume months are still included but faded.

40% Apr 202640% Apr 2026 low40% Sep 202667% last 90 days
0255075100APPROVAL75% all-time40%40%Apr '26May '26Jun '26Jul '26Aug '26Sep '26
Monthly approval, 0–100. Bars show opinion volume. Months with fewer than 3 opinions are left out.
View data table
MonthApprovalOpinions
Apr 202640%5
May 202677%32
Jun 202684%43
Jul 202669%29
Aug 202686%7
Sep 202640%5

What developers say.

Summaries are paraphrased. Each links to the original Hacker News comment.

Praise

4 selected
Cost & pricing

DeepSeek V4 Pro was also cheap enough for their side project.

HN comment · Sep 19, 2026Read on HN ↗
Cost & pricing

Heavy DeepSeek V4 Pro use cost about $24 a month during the promo, far cheaper than running it locally.

HN comment · Sep 15, 2026Read on HN ↗
Cost & pricing

Uses DeepSeek V4 Pro for development, citing its pricing as worth it.

HN comment · Aug 14, 2026Read on HN ↗
Cost & pricing

They now use DeepSeek V4 Pro almost exclusively but worry about planned API price increases.

HN comment · Aug 12, 2026Read on HN ↗

Complaints

4 selected
Reliability

Spent a weekend on DeepSeek V4 Pro for a Linux port; it broke booting and wasted $120.

HN comment · Sep 29, 2026Read on HN ↗
Reliability

DeepSeek V4 Pro got sidetracked, hallucinated problems, and stalled for hours on a simple cleanup.

HN comment · Sep 9, 2026Read on HN ↗
Code quality

The DeepSeek V4 Pro preview kept messing things up, needing 30-60% extra time to fix output.

HN comment · Aug 8, 2026Read on HN ↗
Code quality

Tried DeepSeek V4 Pro and found its code questionable and error-prone, so savings did not speed up progress.

HN comment · Jul 28, 2026Read on HN ↗

Questions developers ask.

Is DeepSeek V4 Pro worth it?
For cost-sensitive work, the data supports yes. It holds 75% approval, and cost and pricing get 55 positive mentions against 3 negative. Users report about $24 a month during a promo and under $100 a month for near-exclusive use.
Is DeepSeek V4 Pro better than GLM 5.2?
The data leans no. GLM 5.2 was named as the preferred alternative 4 times, more than any other model. One user also ranked DeepSeek V4 Pro consistently worse than GLM 5.2 at Max reasoning.
Is DeepSeek V4 Pro reliable for coding?
Reliability is the weak area. It draws 9 negative mentions against 2 positive. Reports mention broken changes that need reverting, hallucinated problems, and 30 to 60% extra time to fix output.

The data behind this page.

Developers give DeepSeek V4 Pro 75% approval across 121 opinions from 85 authors, and 104 of those opinions come from firsthand use. Price drives the positive view: cost and pricing draw 55 positive mentions against 3 negative. Code quality is split at 31 positive and 10 negative, and reliability is the weak spot at 9 negative against 2 positive. Approval over the last 90 days is 67%, below the overall figure, and monthly approval moved from 77% in May to 84% in June to 69% in July, while August and September hold only 7 and 5 opinions, too few to read as a trend.

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.

20,064 comments screened121 about DeepSeek V4 Pro104 first-hand

Read the full methodology →