Product analytics
Asked as “product analytics platform”, and as “product analytics tool”, on behalf of a mid-market B2B company. 58 first choices recorded across the direct, paraphrase, budget and scale prompts, fourteen models each.
Standing · first-choice share
34%
Contested · Mixpanel 24%
34Amplitude24Mixpanel22PostHog19others
34% of first choices, contested.
Since September 2026▲+3Since September 2026: 32% → 35%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.Amplitude held the lead, +3 points on 32%, inside the 11-point floor.
By buyer segment
The same question asked on behalf of a different buyer. Each standing is computed within its segment; they sit side by side and are never added together.
Small business
PostHog50%Clear leader▲+3Since September 2026: 43% → 46%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.Standing →Mid-market
Amplitude34%Contested▲+3Since September 2026: 32% → 35%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.This pageEnterpriseAmplitude56%Clear leader▲+4Since September 2026: 53% → 57%, +4 points. Inside the 11-point floor: within noise. Read over the models both editions asked.Standing →
PostHog50%Clear leader▲+3Since September 2026: 43% → 46%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.Standing →Mid-marketAmplitude34%Contested▲+3Since September 2026: 32% → 35%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.This pageEnterpriseAmplitude56%Clear leader▲+4Since September 2026: 53% → 57%, +4 points. Inside the 11-point floor: within noise. Read over the models both editions asked.Standing →The standing
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all fourteen models, for a mid-market B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrantSince September 2026
01Amplitude34%19%74endorsed leader▲+3Since September 2026: 32% → 35%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.32% → 35%
Amplitude34%19%74endorsed leader▲+3Since September 2026: 32% → 35%, +3 points. Inside the 11-point floor: within noise. Read over the models both editions asked.32% → 35%02
Mixpanel24%14%70accepted challenger▼−4Since September 2026: 28% → 24%, −4 points. Inside the 11-point floor: within noise. Read over the models both editions asked.28% → 24%
Mixpanel24%14%70accepted challenger▼−4Since September 2026: 28% → 24%, −4 points. Inside the 11-point floor: within noise. Read over the models both editions asked.28% → 24%03PostHog22%5%56accepted challenger▲+1Since September 2026: 21% → 22%, +1 point. Inside the 11-point floor: within noise. Read over the models both editions asked.21% → 22%
PostHog22%5%56accepted challenger▲+1Since September 2026: 21% → 22%, +1 point. Inside the 11-point floor: within noise. Read over the models both editions asked.21% → 22%04
Pendo10%13%46accepted challenger▲+1Since September 2026: 11% → 12%, +1 point. Inside the 11-point floor: within noise. Read over the models both editions asked.11% → 12%
Pendo10%13%46accepted challenger▲+1Since September 2026: 11% → 12%, +1 point. Inside the 11-point floor: within noise. Read over the models both editions asked.11% → 12%05
Heap3%24%51accepted challenger▲+4Since September 2026: 0% → 4%, +4 points. Inside the 11-point floor: within noise. Read over the models both editions asked.0% → 4%
Heap3%24%51accepted challenger▲+4Since September 2026: 0% → 4%, +4 points. Inside the 11-point floor: within noise. Read over the models both editions asked.0% → 4%06
Google Analytics 4 Lab in the set2%73%22criticized challenger=heldSince September 2026: 2% → 2%, ±0 points. Inside the 11-point floor: within noise. Read over the models both editions asked.2% → 2%
Google Analytics 4 Lab in the set2%73%22criticized challenger=heldSince September 2026: 2% → 2%, ±0 points. Inside the 11-point floor: within noise. Read over the models both editions asked.2% → 2%Show the one product at 0%, ordered by negative rate
07
Fullstory0%32%19criticized challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 11-point floor: within noise. Read over the models both editions asked.0% → 0%
Fullstory0%32%19criticized challenger=heldSince September 2026: 0% → 0%, ±0 points. Inside the 11-point floor: within noise. Read over the models both editions asked.0% → 0%The floor is 11 points of share, measured: how far the models move a leader on their own when the same questions are asked twice with nothing changed. A larger change is movement; a smaller one can't be told apart from repeat variation, and both are shown. Each index measures its own floor from its own repeat, over its own categories, so the floor differs from one index to the next. Movement is read over the twelve models both editions asked; GPT-6 Luna, Muse Glimmer 30B joined this edition and are in the standing but not yet in the comparison. How the floor is measured
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its product page.
Google Analytics 4 is made by Google, whose model Gemini 3.5 Flash is in the set. On the four share prompts that model made it the first choice zero times of 4; the other thirteen models one time of 52. Lab treatment is defined on the method page; the row is marked, not excluded.
All fifteen head-to-head pages: the top six products, each against each
Recommended versus criticized
Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.
Criticized challengerCriticized default
Negative label rate →
01020304050607Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative50%
Key
01Amplitude34%
Amplitude34%02Mixpanel24%
Mixpanel24%03PostHog22%
PostHog22%04Pendo10%
Pendo10%05Heap3%
Heap3%06Google Analytics 42%
Google Analytics 42%07Fullstory0%
Fullstory0%What they warned about
Eight of fourteen models held their first choice under the paraphrase. Claude Haiku 4.5, Grok 4.20, Mistral Small, Llama 4 Maverick, Kimi K2 and Muse Glimmer 30B changed. A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Google Analytics 473%
16 of 22 labels negative · 13 of 14 models · 7 hard negative
“Avoid GA4 as your main product-analytics system if you care about funnels, retention, feature adoption, or cohort analysis.” Perplexity Sonar, negative prompt
Mixpanel14%
10 of 70 labels negative · 10 of 14 models · 1 hard negative
“"Highest Caution: Mixpanel … When to avoid: If handling sensitive user data or EU users"ate"ated security incidents” Muse Glimmer 30B, negative prompt
Amplitude19%
14 of 74 labels negative · 10 of 14 models
“it is strong for product analytics, but some sources note add-ons ... can require extra line items, and the learning curve can be steep” Perplexity Sonar, negative prompt
Heap24%
12 of 51 labels negative · 9 of 14 models · 2 hard negative
“Post-acquisition decline... Heap and Mixpanel have the most documented customer pain points” Kimi K2, negative prompt
What they cite
Fourteen of the fourteen models returned a source list in this edition: five return the pages they cited and nine the search results they consulted.
Sites the answers cite
73 of 84 answers in this category came back with a source list, from 14 of 14 models: citations where the model returns them, or the search results it consulted. 966 links across 199 sites, every framing counted. Ranked by the number of answers carrying the site or page. 19 of the 252 answers across every segment cited this index's own page for the category; the method page measures whether that reading tilts an answer.
vendor site · Amplitude53 answers · 99 citations · 14 models
vendor site · Userpilot35 answers · 37 citations · 11 models
vendor site · Gainsight32 answers · 35 citations · 11 models
vendor site · Pendo30 answers · 35 citations · 11 models
vendor site · checkthat.ai29 answers · 35 citations · 11 models
vendor site · PostHog26 answers · 44 citations · 11 models
vendor site · CleverX25 answers · 28 citations · 11 models
vendor site · LinkedIn23 answers · 25 citations · 10 models
vendor site · Guideflow23 answers · 24 citations · 8 models
23 answers · 24 citations · 9 models
22 answers · 22 citations · 10 models
vendor site · Mixpanel21 answers · 28 citations · 9 models
Pages the answers cite
The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.
40 answers · 13 models
32 answers · 11 models
25 answers · 9 models
25 answers · 9 models
24 answers · 11 models
23 answers · 9 models
23 answers · 10 models
22 answers · 10 models
16 answers · 8 models
16 answers · 9 models
Search against answers
Each company's standing in the answers beside its site's footprint in Google search, one row a site: the products the models named on it with their shares, and the share they add up to; monthly searches on Google, and DataForSEO's estimate of AI search demand (modeled from search signals, directional, not a count of queries to any assistant), for the most-searched of the company's and its products' names (the name is in each row's hover text); estimated monthly organic visits to the site; and its best position in Google's top ten for “best product analytics platform”, “product analytics platform”. US estimates from DataForSEO and Google's Ads Transparency Center. A small company's site, or a mid-sized company's site for its flagship, is marked company; a product on a large parent's site (Google, Microsoft) has no site figures. A column with no figures for this category is left out, and an empty cell means none were seen, not none exist. Two measurements side by side: neither is read as the cause of the other.
| Company and products | Share | Own site cited | Name searches, Google | AI search demand, est. | Organic visits | Paid search a month, est. | Google ads, last 30 days | Best Google position |
|---|---|---|---|---|---|---|---|---|
| Amplitude amplitude.com | 34% | 418 | 74,000 | 5,691 | 373,273 | $207 | 27 | 2 |
| Mixpanel mixpanel.com | 24% | 467 | 12,100 | 458 | 20,284 | $458 | 11 | |
| PostHog posthog.com | 22% | 420 | 33,100 | 271 | 25,936 | 120 | ||
| Pendo pendo.io | 10% | 409 | 12,100 | 761 | 19,185 | $3,609 | 93 | |
| Contentsquare heap.io Heap 3% | 3% | 150 | 40,500 | 11,454 | 16,541 | 32 | 7 | |
| June june.so | 3% | 13 | 135,000 | 209,401 | 1,167 | |||
| Google analytics.google.com Google Analytics 4 2% | 2% | 473 | 9,900 | 348 | ||||
| Userpilot userpilot.com | 0% | 247 | 1,600 | 10 | 62,084 | 6 | 3 | |
| Fullstory fullstory.com | 0% | 217 | 5,400 | 159 | 141,869 | $348 | 14 | 8 |
| Statsig statsig.com | 0% | 67 | 12,100 | 182 | 353,837 | $176 | 6 |
Google's top ten for the two searches hold 16 sites; 4 of them are among the sites the models cited here (amplitude.com, gainsight.com, learn.g2.com, userpilot.com). The first result for “best product analytics platform” is domo.com.
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.ShowHide
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ShowHideModelDirect“What is the best product analytics platform for a mid-market B2B company?”Paraphrase“Which product analytics tool would you recommend to a mid-sized B2B company?”Comparative“What are the top product analytics platforms and how do they differ?”Budget-constrained“What is the best product analytics platform for a company with a limited budget?”Scale-constrained“We are a 500 person company evaluating a product analytics platform. What should we look at?”Negative“Which product analytics platforms should I avoid or be cautious about?”
Claude Haiku 4.5PendoMixpanelChangedagainst: AmplitudeAmplitudeMixpanel, PostHogagainst: Google Analytics 4no first choiceagainst: Amplitude, Contentsquare, Fullstory, Heap, Mixpanel, Pendo, PostHog
Three alternatives
Amplitude, Heap, MixpanelTwo alternatives
GainsightPX, PendoSix alternatives
Fullstory, Heap, Mixpanel, Pendo, PostHog, Quantum MetricTwo alternatives
Amplitude, UsermavenGPT-5.4 miniAmplitudeAmplitudeHeldAmplitudePostHogagainst: Amplitude, Google Analytics 4no first choicenothing named
Three alternatives
Mixpanel, Pendo, PostHogThree alternatives
Heap, Mixpanel, PostHogFour alternatives
Heap, Mixpanel, Pendo, PostHogOne alternative
MixpanelGemini 3.5 FlashMixpanelMixpanelHeldMixpanelPostHogagainst: AmplitudeMixpanelagainst: Heapagainst: Amplitude, Contentsquare, Fullstory, Google Analytics 4, Heap, Hotjar, LogRocket, Mixpanel, Pendo
Three alternatives
Amplitude, Pendo, PostHogFour alternatives
Amplitude, June, Pendo, PostHogFour alternatives
Amplitude, Heap, Pendo, PostHogTwo alternatives
GrowthBook, MixpanelFour alternatives
Amplitude, Kubit, NetSpring, PostHogPerplexity SonarAmplitudeAmplitudeHeldAmplitudePostHogagainst: Heapno first choiceagainst: Amplitude, Fullstory, Google Analytics 4, Heap, Mixpanel
Four alternatives
Gainsight PX, June, Mixpanel, PendoOne alternative
PendoSix alternatives
Adobe Analytics, Fullstory, Heap, Mixpanel, Pendo, PostHogTwo alternatives
Mixpanel, UXCamGrok 4.20Amplitudeagainst: PostHogPendoChangedno first choiceagainst: Contentsquare, Fullstory, Google Analytics 4Mixpanel, PostHogno first choiceagainst: Amplitude, Google Analytics 4, Heap, Mixpanel
Two alternatives
Mixpanel, PendoFour alternatives
Amplitude, Heap, Mixpanel, PostHogThree alternatives
Amplitude, Google Analytics 4, HotjarMistral SmallAmplitudeJuneChangedAmplitude, MixpanelGoogle Analytics 4, Mixpanelno first choiceagainst: Adobe Analytics, Google Analytics 4, Heap, Hotjar, Pendo
Two alternatives
Heap, PendoThree alternatives
Amplitude, Mixpanel, PendoThree alternatives
Heap, Pendo, PostHogThree alternatives
Amplitude, Heap, HotjarDeepSeek V4 FlashAmplitudeagainst: LogRocketAmplitudeHeldagainst: Fullstory, Gainsight PX, Heap, MixpanelAmplitude, MixpanelPostHogagainst: Google Analytics 4no first choicenothing named
Two alternatives
Mixpanel, PostHogOne alternative
PostHogFour alternatives
Fullstory, Heap, Pendo, PostHogOne alternative
AmplitudeLlama 4 MaverickAmplitude, HeapHeapChangedno first choiceMixpanelAmplitude, Pendoagainst: Adobe Analytics, Google Analytics 4
Three alternatives
Amplitude, June, UserpilotThree alternatives
Amplitude, Contentsquare, UserpilotQwen 3.7 FlashAmplitudeAmplitudeHeldAmplitude, Mixpanel, PostHogagainst: Hotjar, Microsoft ClarityMixpanel, PostHogagainst: AmplitudeAmplitude, Mixpanel, Pendoagainst: Amplitude, Google Analytics 4, Heap, Mixpanel
Four alternatives
Heap, Mixpanel, Pendo, PostHogTwo alternatives
Pendo, PostHogTwo alternatives
Heap, PendoOne alternative
MatomoThree alternatives
Fullstory, Heap, PostHogKimi K2Mixpanel, PostHogPostHogChangedAmplitude, Mixpanelagainst: Google Analytics 4PostHogagainst: Amplitude, Heapno first choiceagainst: Contentsquare, Google Analytics 4, Heap, Hotjar, Mixpanel, Pendo
Two alternatives
Amplitude, HeapOne alternative
AmplitudeThree alternatives
Fullstory, Heap, PendoTwo alternatives
Google Analytics 4, MixpanelGLM 4.7 FlashXAmplitudeagainst: PendoAmplitudeHeldAmplitude, Mixpanelagainst: Google Analytics 4PostHogagainst: Amplitudeno first choiceagainst: Amplitude, Google Analytics 4, Heap, Mixpanel, Pendo
Three alternatives
Heap, Mixpanel, PostHogOne alternative
MixpanelFive alternatives
Fullstory, Glassbox, Heap, Pendo, UXCamOne alternative
MixpanelMiniMax M2.5Amplitude, PendoAmplitude, PendoHeldAmplitudeAmplitude, Mixpanelno first choicenothing named
Three alternatives
Heap, Mixpanel, PostHogTwo alternatives
June, MixpanelThree alternatives
Heap, Mixpanel, PostHogTwo alternatives
Plausible, PostHogGPT-6 LunaMixpanelMixpanelHeldno first choicePostHogno first choiceagainst: Amplitude, Google Analytics 4, Mixpanel, PostHog
Two alternatives
Amplitude, PostHogTwo alternatives
Amplitude, PostHogFive alternatives
Amplitude, Heap, Mixpanel, Pendo, PostHogTwo alternatives
Amplitude, MixpanelMuse Glimmer 30BAmplitudeAmplitude, JuneChangedAmplitudePostHogno first choiceagainst: Amplitude, Fullstory, Google Analytics 4, Heap, by Contentsquare, Mixpanel
Five alternatives
Heap, June, Mixpanel, Pendo, PostHogTwo alternatives
Mixpanel, PostHogFour alternatives
Heap, Mixpanel, Pendo, StatsigTwo alternatives
Amplitude, OpenPanelTwo alternatives
Amplitude, PendoBold is the first choiceAlternatives are counted; the count opens them.What the answer argued against
The record
One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency.
Eighty-four rows: every prompt, every model, every answer.
| Prompt | Model | Version string | Time (UTC) | Searched | Sources | Latency |
|---|---|---|---|---|---|---|
| Direct recommendation | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-01 09:42 | yes | 9 | 8 s |
| Direct recommendation | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-01 07:40 | yes | 5 | 4 s |
| Direct recommendation | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-01 11:23 | yes | 22 | 28 s |
| Direct recommendation | Perplexity Sonar | sonar | 2026-10-01 10:49 | yes | 19 | 2 s |
| Direct recommendation | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-10-01 13:04 | yes | 23 | 10 s |
| Direct recommendation | Mistral Small | mistral/mistral-small via mistral | 2026-10-01 12:26 | yes | 9 | 4 s |
| Direct recommendation | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-01 09:55 | yes | 17 | 29 s |
| Direct recommendation | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-01 08:23 | yes | 5 | 1 s |
| Direct recommendation | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-01 09:49 | no | 0 | 37 s |
| Direct recommendation | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-01 08:08 | yes | 13 | 20 s |
| Direct recommendation | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-01 09:04 | yes | 14 | 39 s |
| Direct recommendation | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-01 13:53 | yes | 15 | 24 s |
| Direct recommendation | GPT-6 Luna | gpt-6-luna | 2026-10-01 12:12 | yes | 3 | 16 s |
| Direct recommendation | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-01 13:25 | yes | 14 | 25 s |
| Paraphrase | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-01 08:48 | yes | 9 | 8 s |
| Paraphrase | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-01 09:51 | yes | 5 | 4 s |
| Paraphrase | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-01 13:09 | yes | 9 | 21 s |
| Paraphrase | Perplexity Sonar | sonar | 2026-10-01 09:39 | yes | 25 | 3 s |
| Paraphrase | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-10-01 11:14 | yes | 22 | 8 s |
| Paraphrase | Mistral Small | mistral/mistral-small via mistral | 2026-10-01 09:15 | yes | 10 | 7 s |
| Paraphrase | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-01 11:14 | yes | 23 | 36 s |
| Paraphrase | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-01 10:48 | yes | 5 | 1 s |
| Paraphrase | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-01 09:12 | no | 0 | 24 s |
| Paraphrase | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-01 07:54 | yes | 25 | 31 s |
| Paraphrase | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-01 12:41 | yes | 15 | 68 s |
| Paraphrase | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-01 07:40 | yes | 5 | 21 s |
| Paraphrase | GPT-6 Luna | gpt-6-luna | 2026-10-01 09:08 | yes | 2 | 13 s |
| Paraphrase | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-01 12:19 | yes | 14 | 18 s |
| Comparative | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-01 12:42 | yes | 9 | 9 s |
| Comparative | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-01 11:17 | yes | 5 | 7 s |
| Comparative | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-01 14:14 | yes | 20 | 27 s |
| Comparative | Perplexity Sonar | sonar | 2026-10-01 12:07 | yes | 15 | 5 s |
| Comparative | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-10-01 11:37 | yes | 19 | 9 s |
| Comparative | Mistral Small | mistral/mistral-small via mistral | 2026-10-01 12:08 | yes | 10 | 7 s |
| Comparative | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-01 13:48 | yes | 21 | 52 s |
| Comparative | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-01 08:21 | yes | 5 | 1 s |
| Comparative | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-01 11:42 | yes | 10 | 28 s |
| Comparative | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-01 11:37 | yes | 15 | 29 s |
| Comparative | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-01 10:02 | yes | 17 | 174 s |
| Comparative | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-01 12:36 | yes | 13 | 19 s |
| Comparative | GPT-6 Luna | gpt-6-luna | 2026-10-01 11:20 | yes | 6 | 19 s |
| Comparative | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-01 07:27 | yes | 11 | 26 s |
| Budget constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-01 12:32 | yes | 9 | 8 s |
| Budget constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-01 09:41 | no | 0 | 4 s |
| Budget constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-01 09:10 | yes | 12 | 18 s |
| Budget constrained | Perplexity Sonar | sonar | 2026-10-01 09:12 | yes | 18 | 2 s |
| Budget constrained | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-10-01 13:45 | yes | 19 | 7 s |
| Budget constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-01 11:32 | yes | 10 | 5 s |
| Budget constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-01 14:08 | yes | 20 | 25 s |
| Budget constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-01 07:25 | yes | 5 | 2 s |
| Budget constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-01 08:10 | yes | 10 | 30 s |
| Budget constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-01 10:27 | yes | 15 | 25 s |
| Budget constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-01 12:03 | yes | 25 | 132 s |
| Budget constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-01 11:55 | no | 0 | 7 s |
| Budget constrained | GPT-6 Luna | gpt-6-luna | 2026-10-01 13:12 | yes | 3 | 13 s |
| Budget constrained | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-01 11:55 | yes | 14 | 21 s |
| Scale constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-01 08:06 | no | 0 | 7 s |
| Scale constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-01 14:14 | no | 0 | 7 s |
| Scale constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-01 09:43 | no | 0 | 19 s |
| Scale constrained | Perplexity Sonar | sonar | 2026-10-01 07:55 | yes | 18 | 5 s |
| Scale constrained | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-10-01 08:12 | yes | 16 | 6 s |
| Scale constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-01 08:10 | no | 0 | 8 s |
| Scale constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-01 10:58 | yes | 23 | 38 s |
| Scale constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-01 09:46 | yes | 5 | 2 s |
| Scale constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-01 13:31 | no | 0 | 30 s |
| Scale constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-01 09:29 | yes | 10 | 19 s |
| Scale constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-01 09:32 | yes | 15 | 107 s |
| Scale constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-01 08:22 | no | 0 | 8 s |
| Scale constrained | GPT-6 Luna | gpt-6-luna | 2026-10-01 08:00 | yes | 2 | 28 s |
| Scale constrained | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-01 13:24 | yes | 14 | 25 s |
| Negative framing | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-01 13:23 | yes | 16 | 10 s |
| Negative framing | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-01 10:04 | yes | 5 | 8 s |
| Negative framing | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-01 13:32 | yes | 13 | 30 s |
| Negative framing | Perplexity Sonar | sonar | 2026-10-01 10:11 | yes | 20 | 5 s |
| Negative framing | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-10-01 12:29 | yes | 25 | 8 s |
| Negative framing | Mistral Small | mistral/mistral-small via mistral | 2026-10-01 08:24 | yes | 5 | 6 s |
| Negative framing | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-01 13:29 | yes | 25 | 12 s |
| Negative framing | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-01 12:33 | yes | 5 | 2 s |
| Negative framing | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-01 12:12 | yes | 5 | 34 s |
| Negative framing | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-01 08:40 | yes | 25 | 27 s |
| Negative framing | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-01 09:28 | yes | 22 | 94 s |
| Negative framing | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-01 13:24 | no | 0 | 10 s |
| Negative framing | GPT-6 Luna | gpt-6-luna | 2026-10-01 12:03 | yes | 4 | 19 s |
| Negative framing | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-01 13:42 | yes | 20 | 22 s |
Noise floor in this category
Flips between the edition run and its calibration repeat. Six prompts per model is a small sample; the index-wide floor is the number to trust.
ShowHide
Claude Haiku 4.5
4 of 4 flipped
GPT-5.4 mini
0 of 4 flipped
Gemini 3.5 Flash
5 of 6 flipped
Perplexity Sonar
2 of 4 flipped
Grok 4.20
5 of 6 flipped
Mistral Small
2 of 4 flipped
DeepSeek V4 Flash
2 of 4 flipped
Llama 4 Maverick
2 of 4 flipped
Qwen 3.7 Flash
4 of 5 flipped
Kimi K2
4 of 5 flipped
GLM 4.7 FlashX
2 of 5 flipped
MiniMax M2.5
4 of 4 flipped
GPT-6 Luna
3 of 4 flipped
Muse Glimmer 30B
2 of 4 flipped
Normalization in this category
Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.
ShowHide
Category-scoped readings
Amplitude read as Amplitude
Unresolved, counted raw
Heap, by Contentsquare
Klaritics
Whatfix Product Analytics
Discontinued, still offered
No shut-down product was recommended here.