A.X-K2 wins more areas, narrowly.K2 Horizon 7B is better at agents.
Scores updated · 13 tests both models report · How we compare
Where each one wins
Tests won in each of the five areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking30A.X-K23 of 3 tests
- FactsGetting facts right instead of making them up10A.X-K21 of 2 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own03K2 Horizon 7B3 of 3 tests
- CodingWriting and fixing software11Even1 each
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
Long documents rests on a single test.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where A.X-K2 pulls ahead
- Code for real scientific research problemsSciCode+13.5points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+11.4points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+10.4points ahead
Where K2 Horizon 7B pulls ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+11.5points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+5.2points ahead
- Real work tasks from 44 professionsGDPVal+3points ahead
Every test, side by side
All 13 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- SciCodeA.X-K2 by 13.54127.5+13.5
- Terminal-Bench 2.1K2 Horizon 7B by 5.23944.2+5.2
AgentsK2 Horizon 7B
- τ-Bench V3 · BankingK2 Horizon 7B by 11.516.127.6+11.5
- GDPValK2 Horizon 7B by 323.226.2+3
- Terminal-Bench 4.0K2 Horizon 7B by 101+1
ReasoningA.X-K2
- Humanity's Last ExamA.X-K2 by 11.429.618.2+11.4
- GPQA DiamondA.X-K2 by 10.485.775.3+10.4
- CritPtA.X-K2 by 8.68.90.3+8.6
FactsA.X-K2
- AA-Omniscience · AccuracyA.X-K2 by 5.818.612.8+5.8
- AA-Omniscience · Non-hallucinationtie6766.4tie
Other results2 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA-OmniscienceA.X-K2 by 8.4-8.2-16.6+8.4
- AA Intelligencetie21.520.6tie
Questions people ask
Which is better, A.X-K2 or K2 Horizon 7B?
A.X-K2 wins two of the five areas where both have results: reasoning and facts. K2 Horizon 7B wins agents. They are level on coding and long documents.
Which is better for coding?
Neither. They win 1 coding test each of the 2 both models report.
How do you compare the two?
We use the 13 benchmark tests both models have published scores on. The verdict counts the 11 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 2 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.