20% off all GPT 🎉 30% off DeepSeekLearn more

AI Model Usage Leaderboard

See what developers use on Ofox and how those choices change over time.

Anonymized, aggregated usageData through August 30, 2026

Usage over time

ModelsLabsApps
0%25%50%75%100%W28W29W30W31W32W33W34W35

Top 20 by usage

8-week summary
01
AnthropicClaude Opus 4.7
13.6%
-7.3pp
02
AnthropicClaude Opus 4.8
11.0%
-30.7pp
03
OpenAIGPT 5.6 Luna
11.0%
+18.7pp
04
AnthropicClaude Sonnet 4.6
10.4%
-6.2pp
05
AnthropicClaude Sonnet 5
7.7%
-1.4pp
06
OpenAIGPT 5.6 Sol
5.9%
+10.4pp
07
DeepSeek
DeepSeek V4 Flash
4.3%
-2.8pp
08
ZhipuGLM 5.2
4.1%
-3.9pp
09
AnthropicClaude Opus 5
3.6%
+3.8pp
10
DeepSeek
DeepSeek V4 Pro
3.2%
-5.2pp
11
OpenAIGPT 5.5
3.0%
-4.3pp
12
OpenAIGPT 5.6 Terra
2.2%
+2.3pp
13
ZhipuGLM 5.3 Flash
1.8%
+13.8pp
14
QwenQwen3.8 Max
1.1%
+2.2pp
15
AnthropicClaude Fable 5
1.0%
-0.5pp
16
OpenAIGPT 5.4
0.9%
-1.1pp
17
MoonshotAIKimi K3
0.8%
+1.3pp
18
AnthropicClaude Opus 4.6
0.8%
-0.7pp
19
AnthropicClaude Haiku 4.5
0.7%
-0.5pp
20
•••
Other
12.9%
+12.1pp

How to read this board

Use the sample and coverage dates as context, then compare models for your task and budget.

Data through

Sample
Usage shares within Ofox, not global market share or a quality ranking. Large callers can materially affect these shares.
Coverage
Grouped by UTC date, using only weeks with complete date coverage. The chart covers historical periods, not real-time usage.
Metrics
Token share measures token usage; request share measures call counts. Calls billed by seconds or other units without tokens are not represented by token share.
App labels
Based on request-client information, not unique users. Unknown or non-public client labels are grouped into Other.

Popularity helps you compare. Fit helps you decide.

Balance capability, price, and use case to find the model that works for you.

Find a model for your use caseBrowse all models