AI discovery pulse
What’s Trending in AI
A fast view of the products, models, stories, companies and comparisons attracting the most attention across the AI Orbit dataset.
320Published tools
134Active models
0News / 24h
68Comparisons
Leading model performance
Model benchmarks Models to Watch
Active and preview models surfaced by benchmark score and release recency.
#1
Anthropic
Claude Sonnet 5
Active
Benchmark85.2Context1,000,000Input / 1M$2.00
#2
OpenAI
GPT-6 Astra
Active
Benchmark80.6Context1050000Input / 1M$10.00
#3
DeepSeek
DeepSeek-R1
Active
Benchmark76.7Context128,000Input / 1MNot verified
#4
OpenAI
GPT-5
Active
Benchmark74.7Context400,000Input / 1M$1.25
#5
OpenAI
o3
Active
Benchmark69.8Context200,000Input / 1M$2.00
#6
OpenAI
GPT-5 mini
Active
Benchmark69.5Context400,000Input / 1M$0.25
#7
Google DeepMind
Gemini 2.5 Pro
Active
Benchmark69.4Context1,048,576Input / 1M$1.25
#8
OpenAI
GPT-5.6
Active
Benchmark68.6Context1,050,000Input / 1M$4.00
#9
OpenAI
GPT-5.6 Sol
Active
Benchmark68.6Context1,050,000Input / 1M$4.00
#10
Anthropic
Claude Fable 5.1
Active
Benchmark67.3Context1000000Input / 1M$10.00
#11
Google DeepMind
Gemini 3.8 Flash Cyber
Active
Benchmark66.7Context—Input / 1M$0.75
#12
OpenAI
o4-mini
Active
Benchmark66.5Context200,000Input / 1M$1.10
What “trending” means here
For AI tools, AI Orbit now uses first-party rolling 30-day activity: unique profile views, search-result clicks and direct tool searches. Other sections continue to use their verified internal signals such as benchmark performance, editorial news importance, ecosystem activity and comparison views. No external or fabricated traffic percentage is used.