Anuma
What you can do
ChatCreateBuildSolve
Core Features
Council ModePortable ContextUnified MemoryMultiple AI ModelsPrivate AI
Work & Career
ProfessionalsDevelopersEntrepreneursJob SeekersFreelancersSmall Business OwnersParents
Wellness & Health
Health & FitnessMental WellnessFood & Nutrition
Students & Learning
StudentsResearchersLanguage LearnersTeachers & Educators
Creative & Content
WritersDesignersContent CreatorsMusicians
Finance & Legal
Personal FinanceLegal & ContractsInvestment ClubsReal Estate
AboutCareersBrandingContact UsAffiliate
Help CenterWhat's NewBlogResearchFAQAI Prompt LibraryHow Memory WorksOpen-Source LLMsClosed-Source LLMsPrivate by Default
Pricing
Get the appTry Anuma
Anuma
What you can do
ChatCreateBuildSolve
Core Features
Council ModePortable ContextUnified MemoryMultiple AI ModelsPrivate AI
Work & Career
ProfessionalsDevelopersEntrepreneursJob SeekersFreelancersSmall Business OwnersParents
Wellness & Health
Health & FitnessMental WellnessFood & Nutrition
Students & Learning
StudentsResearchersLanguage LearnersTeachers & Educators
Creative & Content
WritersDesignersContent CreatorsMusicians
Finance & Legal
Personal FinanceLegal & ContractsInvestment ClubsReal Estate
AboutCareersBrandingContact UsAffiliate
Help CenterWhat's NewBlogResearchFAQAI Prompt LibraryHow Memory WorksOpen-Source LLMsClosed-Source LLMsPrivate by Default
Pricing
Get the appTry Anuma
The Consumer Model IndexUpdated July 10, 2026

Consumer AI Model Rankings

Ranked from how people really use AI through Anuma — the models they pick, switch to, and stick with across every major provider, plus independent research on the shifts.

Top models by deliberate picks · past 30 days
#ModelShare of picks%Δ 90d
1gpt-5.421.6▼ 2.8
2qwen-3.6-max18.4▲ 14.2
3grok-4.313.9▲ 3.1
4kimi-k2.68.5▲ 5.0
5claude-sonnet-4.68.5▲ 0.7
6claude-opus-4.83.8▲ 2.3
7gemini-3.1-pro3.7▼ 0.7
8gpt-5.53.6▼ 1.0
9gemini-3.5-flash2.6NEW
10qwen-3.6-plus2.3▲ 0.8
11minimax-m2.71.8NEW
12claude-sonnet-51.8NEW
anuma.ai8,980 deliberate picks · frontend-sampled
13.5B
tokens processed · 90d
104K
conversations · 90d
37
models tracked
14
providers
24%
of users pick their own model
Flows

Where users defect

Of 1,258 users who tried more than one model, we tracked 4,471 switches between consecutive messages. The signature Anuma metric: not which model scores highest on a benchmark, but which one a real person leaves — and where they go next.

Net switcher flow
Gained − lost · 90 days
gemini-3.1-pro
+31
grok-4.3
+28
kimi-k2.6
+21
gemini-3.5-flash
+14
qwen-3.6-plus
+12
claude-sonnet-4.6
-3
claude-opus-4.7
-18
gpt-5.4
-51
kimi-k2.5
-53
grok-4.1-fast-reasoning
-60
anuma.aiConsumer Model Index
Net = users switching to a model minus those switching away. Small samples — directional, not definitive.
Top switch routes
Most common model → model switches · 90 days
gpt-5.4→claude-sonnet-4.6
117 sw
gpt-5.4→grok-4.3
93 sw
claude-sonnet-4.6→gpt-5.4
93 sw
claude-opus-4.7→gpt-5.5
92 sw
gpt-5.4→gpt-5.5
89 sw
gpt-5.5→gpt-5.4
87 sw
gpt-5.5→claude-opus-4.7
76 sw
grok-4.3→gpt-5.4
76 sw
anuma.aiConsumer Model Index
1,258 users who used ≥2 models · 4,471 switches tracked.
Usage

How people use Anuma

Two honest layers, both read from metadata — never a word of the prompt. First, the feature people pick in the composer. Then, the tools the assistant reaches for to answer — web search, weather, and the rest — which fire silently inside a normal chat.

Features people choose
Share of messages by feature · trailing 90 days
Standard chat80.1%
Top pick · GPT-5.4
Image generation14%
Auto-routed pipeline
Video generation2.7%
Council2%
anuma.aiConsumer Model Index
Cells sized by share of all messages. The feature is recorded as the tool a person invoked — never by reading the prompt. “Top pick” is the leading model when users override auto; most messages in every mode are auto-routed (see hover). Image, video and audio route to fixed generation pipelines.

On Modes, Anuma is still ~80% plain chat; the tell is in the tail — people reach for Opus 4.7 in Deep research and GPT-5.2 in Council. Flip to Tools and the picture the modes can't show appears: web search is the single most-invoked tool — ~48% of all tool calls — even though it fires silently inside an ordinary chat.

Different denominators: Modes is a share of all messages; Tools is a share of the ~6% of messages that fire a tool — so the two aren't meant to be read against each other.

Reports

Analysis & deep-dives

The research behind the numbers — data insights on what's moving, technical write-ups on how Anuma works, and perspective pieces on where the model market is going.

FeaturedTechnicalJuly 10, 2026

Inside Anuma's PII redaction

As Europe argues over scanning private messages, how Anuma strips personal data from prompts on-device — and what our own usage data shows about privacy in practice.

Read the analysis →
DataJuly 7, 2026

The Qwen Surge — and sudden sunset

Qwen-3.6-Max rocketed from 4% of deliberate picks to the #2 spot in a month — then fell to near-zero the next week when it was retired. A case study in how fast consumer model preference moves.

Read →
DataJune 28, 2026

Where users defect: the models people quietly leave

1,258 users, 4,471 switches — the models losing the most users, and who picks them up.

Coming soon
PerspectiveJune 24, 2026

The case for model-agnostic AI

Why betting on one model is the wrong call — and what routing across all of them unlocks.

Coming soon
TechnicalJune 19, 2026

Council Mode, explained

Ask several models the same question and compare — how it works, and how people actually use it.

Coming soon
Method

How the index is built

Two data streams — one for what people choose, one for what the platform serves — plus five rules that bound what the numbers mean.

Rankings & switchingHuman sends · client-side
Tokens & volumeServer-side inference log
01Deliberate choices only
The auto router — now most messages — is resolved server-side, so it is excluded. This measures choice, not routing.
02Human, not agents
100% human sends from the app. Agents, background jobs and embeddings are excluded from rankings — but still counted in platform token totals.
03Usage, not quality
We report what people chose, not benchmark scores. Plan and credit gating shapes the mix.
04Names canonicalized
Hundreds of raw model strings collapse to real models before counting. Δ 90d compares current share to the 90-day baseline.
05Population
A wallet-gated, mobile-heavy base — directional and early, not a general-market sample.

What you can do

  • Chat
  • Create
  • Build
  • Solve

Core features

  • Unified Memory
  • Multiple AI Models
  • Council Mode
  • Private by Default

Solutions

  • Work & Career
  • Student & learning
  • Creative & Content
  • Wellness & Health
  • Finance & Legal

Company

  • About
  • Careers
  • Branding
  • Contact us
  • Affiliate

Resources

  • Help Center
  • Blog
  • FAQ
  • Prompt Library
  • How memory works
  • Open-Source LLM
  • Closed-Source LLM
Download for iOSDownload for Android
Anuma
Powered by
© 2026 Anuma, Inc. All rights reserved.|Privacy|Terms|Cookie Policy||