Institute of Foundation Models Released September 2026
How K2 Horizon 375B A23B compares with the other 86 ranked language models.
Better for simple tasks. Ranked #70 of 87 on the overall intelligence score, ahead of 20% of models.
Fast writer. Writes about 563 words per second, faster than 100% of models.
Short wait before answering. You wait about 4.1 sec before the answer starts.
Good for
Not ideal for
Intelligence
18
overall score
#70 of 87typical 30
Coding
–
coding score
typical –
Speed
750
tokens per second
#1 of 87typical 91
Wait
4.1 sec
before it replies
#15 of 77typical 15s
Price
–
per 1M tokens
typical $2.00
Scores on individual tests that make up the intelligence score. The faint line marks the best score of any model.
Humanity's Last Exam
Very hard expert questions across many subjects
GPQA Diamond
PhD-level biology, physics and chemistry questions
SciCode
Writing code to solve science research problems
Long documents
Finding and reasoning over facts in very long texts
Terminal-Bench
Doing real tasks on a computer by itself, like an agent
Customer support
Helping bank customers as a support agent using tools
This model next to the models closest to it. Click another bar to open that model.
Through the API, at the listed price. Chat apps usually charge a monthly plan instead.
1,000 chat replies
a question and a ~350-word answer, 1,000 times
Summarise a 100-page report
about 50,000 words in, a one-page summary out
Write a 2,000-word article
a short brief in, a long article out
Input, per 1M tokens
$0.00
Output, per 1M tokens
$0.00
Median times measured through the API.
1 token is about ¾ of a word. Speeds change with server load, so treat these as typical, not guaranteed.
Other companies' models with the closest intelligence score.
Benchmark data from Artificial Analysis, an independent benchmarking company. Rankings refresh every few hours.
kitio.ai