AI Tools and Software Sentiment Report: Week of August 25, 2025
August 26, 2025
This week's sentiment report covers aggregated public online discussion of tracked AI tools and software, drawing on mentions analyzed for August 18, 2025 through August 25, 2025. Across the eligible products, we processed a total of 2,212 relevant mentions, and the picture that emerges is one of steady leaders at the top, a handful of sharp movers, and a complaint conversation dominated by reliability concerns.
Everything here reflects the tone of community sentiment in public online discussion, not a verdict on any product's quality. The reporting period running from August 18, 2025 through August 25, 2025 forms the latest complete week in our digest, and where we cite a score we also cite the sample size behind it so readers can weigh how thin or thick each read is.
Todoist tops the leaderboard with a Pulse Score of 70, based on 26 mentions over the period. It is followed by Obsidian at 68, based on 73 mentions, and Aider at 64, based on 40 mentions. Claude sits at 63, based on 20 mentions, and Mistral rounds out the top five at 59, based on 46 mentions. The spread here is narrow, with just 11 points separating the top and fifth entries.
Sample sizes vary widely across these leaders, which matters for how confidently we read each score. Obsidian's 73 mentions give it the deepest base among the leaders, while Claude's 20 mentions and Todoist's 26 mentions rest on thinner samples. A high score on a smaller sample can move more easily week to week, so the leaderboard is best read as a snapshot of tone rather than a fixed ranking.
Category View
At the category level, most movements over the period were modest. Business led the categories, moving from 55 to 57, and security posted the largest positive shift, rising from 40 to 46. Communication edged up from 38 to 41, and marketing improved from 36 to 38. On the softer side, e-commerce slipped from 50 to 46, design eased from 48 to 46, project-management moved from 50 to 48, and ai-chat cooled slightly from 56 to 54. Finance and education held flat at 35 and 39 respectively. The overall read is of a market where category tone shifted only a few points in either direction, with security the clearest gainer this period.
Biggest Movers
NordVPN was the biggest riser, moving from 44 to 58 across the period, based on 10 mentions. Its latest praise themes leaned on security praise (8 mentions), fair pricing (6 mentions), and strong features (5 mentions), a mix that points to positive discussion of value and protection. That said, its complaint themes still include bugs (12 mentions), reliability (9 mentions), and privacy concerns (6 mentions), so the improvement in tone sits alongside an active critical thread. With only 10 mentions, this is a thin sample that can swing sharply.
Foxit rose from 38 to 44, based on 10 mentions. Praise centered on strong features (10 mentions), comparisons to rivals (6 mentions), and performance (5 mentions), while complaints noted bugs (5 mentions) and UI frustrations (3 mentions). The gain here is more measured than NordVPN's, and again rests on a small base.
Brevo also climbed from 38 to 44, based on 19 mentions. Its praise themes were strong features (7 mentions), fair pricing (6 mentions), and easy to use (5 mentions), suggesting community discussion valued its accessibility and cost. Complaints flagged bugs (7 mentions), reliability (4 mentions), and lacking integrations (3 mentions), a reminder that even a riser carries a mixed conversation.
Confluence was the biggest faller, dropping from 44 to 32, based on 16 mentions. The complaint side was heavy, led by bugs (111 mentions), reliability (61 mentions), and missing features (58 mentions), which together explain much of the negative shift. Praise did exist, with good integrations (17 mentions) and strong features (14 mentions), but the volume of complaint discussion clearly weighed on the tone.
Perplexity fell from 51 to 42, based on 31 mentions. Its complaint themes were unusual in that pricing too high (21 mentions) topped the list, followed by comparisons to rivals (18 mentions) and privacy concerns (13 mentions). Praise for strong features (19 mentions) and AI quality (18 mentions) was present, but the cost and comparison discussion appears to have pulled sentiment lower.
Bun slipped from 41 to 33, based on 21 mentions. Complaints were dominated by bugs (11 mentions), reliability (8 mentions), and lacking integrations (5 mentions), while praise was limited to easy to use (4 mentions), strong features (4 mentions), and performance (2 mentions). The imbalance between a small praise base and a broader complaint set aligns with the downward move.
Spotlight: Todoist
Todoist, a business category product, holds the top Pulse Score this period at 70, based on 26 mentions over the period. Its weekly series shows a score of 71 on August 11, 2025 and 70 on August 25, 2025, a nearly flat line that suggests community sentiment has been stable rather than volatile at the top of the board.
The theme breakdown helps explain the position. Praise discussion leaned on easy to use (9 mentions) and strong features (6 mentions), with feature requests (2 mentions) also appearing, a pattern that reads as users engaged enough to ask for more. On the complaint side, bugs (8 mentions), reliability (6 mentions), and missing features (4 mentions) show the conversation is not uniformly positive.
What stands out is that Todoist's praise and complaint volumes are both small and fairly balanced, yet the tone still lands at the top of the leaderboard. On a 26 mention base, that read should be treated as a stable but modest-sample signal, and worth watching for whether the ease-of-use praise continues to outweigh the bug and reliability mentions in future weeks.
Themes Driving the Conversation
On the praise side, strong features led the latest week with 1,562 mentions, followed by AI quality with 862 mentions and easy to use with 621 mentions. Good integrations drew 369 mentions and comparisons to rivals drew 250 mentions. The strong-features theme shows up across a wide range of products, from Claude (229 mentions) and ChatGPT (183 mentions) to Stripe (84 mentions), which is why it dominates the praise tally. AI quality is concentrated in the ai-chat and related products, with Claude (234 mentions) and Gemini (149 mentions) among the biggest drivers.
On the complaint side, bugs was by far the most-discussed theme with 2,200 mentions, ahead of reliability with 1,383 mentions and missing features with 521 mentions. AI quality also appeared as a complaint with 403 mentions, and comparisons to rivals drew 262 mentions. The bug and reliability themes cut across many products with large complaint volumes, including ArgoCD (231 and 172 mentions), Aspire (224 and 151 mentions), Stripe (188 and 158 mentions), and Vercel (170 and 151 mentions). Notably, AI quality appears on both the praise and complaint lists, a sign that community sentiment about model output is genuinely split depending on the product and the use case.
Watchlist
A large number of tracked products did not clear the threshold of 10 relevant mentions in the period and are therefore excluded from the rankings. This is a matter of insufficient discussion volume, not a judgment on quality. Among the closer-to-threshold names, DALL-E recorded 9 relevant mentions, Veo and Bitbucket each recorded 8 relevant mentions, Anyscale also recorded 8 relevant mentions, and Leonardo AI, ClickUp, Monday.com, Pipedrive, and Pangram Labs each recorded 7 relevant mentions.
Others sat lower still, including HeyGen with 6 relevant mentions, Wrike, Asana, Devin, Jasper, Apollo.io, Braintrust, and Loom each with 5 relevant mentions, and Thinkific, Minimus, Namecheap, and Thinkific-adjacent education tools with 4 relevant mentions. Several well-known names such as Canva (2 mentions), HubSpot (2 mentions), and Linear (2 mentions) drew only a handful of mentions this period. When discussion volume is this thin, any single post can distort a score, which is exactly why we hold these products out of the ranked view until their sample stabilizes.
What To Watch Next Week
First, watch whether security holds its gains. The category rose from 40 to 46 over the period, and NordVPN's move from 44 to 58, based on 10 mentions, was a large part of that story. Because that read rests on a thin sample, it is worth monitoring whether the tone persists or reverts as more mentions arrive.
Second, watch the fallers for stabilization. Confluence dropped to 32, based on 16 mentions, and Perplexity fell to 42, based on 31 mentions, both carrying heavy complaint discussion around bugs, reliability, pricing, and comparisons to rivals. Whether those complaint themes keep dominating will shape where these products land next week.
Third, watch the leaders at the top. Todoist at 70, Obsidian at 68, and Aider at 64 are separated by only a few points, and Todoist's line has been nearly flat. Any shift in the balance between their ease-of-use praise and their bug and reliability complaints could reorder the top of the board.
Frequently Asked Questions
Which product had the highest Pulse Score this period?
Todoist had the highest Pulse Score at 70, based on 26 mentions over the period, narrowly ahead of Obsidian at 68, based on 73 mentions.
Which product moved the most this period?
NordVPN was the biggest riser, climbing from 44 to 58 across the period, based on 10 mentions. On the downside, Confluence fell the most, dropping from 44 to 32, based on 16 mentions.
How many mentions were analyzed this period?
We analyzed a total of 2,212 relevant mentions across 54 products that were eligible for rankings, out of 2,235 products tracked.
What was the most-discussed complaint theme?
Bugs was the most-discussed complaint theme in the latest week with 2,200 mentions, ahead of reliability with 1,383 mentions and missing features with 521 mentions.
About This Data
Pulse Scores summarize the tone of public online discussion on a 0-100 scale and reflect community sentiment, not a verdict on a product's quality or a recommendation. We report on complete calendar weeks only, and products with fewer than 10 relevant mentions in the period are excluded from rankings to avoid unstable reads on thin samples.
Public discussion is collected from Hacker News, Stack Exchange, GitHub, Bluesky, the Apple App Store, and YouTube. Automated sentiment analysis can misread sarcasm, jokes, or niche context, mention volumes vary widely between products, and scores can move week to week. Any company that wants to respond is welcome to reach out. For how scores are calculated, see our methodology.
