GLM-4.5 vs Kimi K2 Instruct
Wins 2 of 6 areas
Reasoning · Following instructions
Wins 1 of 6 areas
Coding
GLM-4.5 wins more areas, narrowly.Kimi K2 Instruct is better at coding.
Scores updated · 23 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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 thinking20GLM-4.52 of 3 tests · 1 tie
- Following instructionsDoing exactly what it is asked10GLM-4.51 of 1 test
- CodingWriting and fixing software12Kimi K2 Instruct2 of 3 tests
- AgentsCarrying out multi-step tasks on its own11Even1 each
- FactsGetting facts right instead of making them up00Even0 each · 2 ties
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
GLM-4.5 costs 10% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where GLM-4.5 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+12.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+6.6points ahead
- Code for real scientific research problemsSciCode+4.1points ahead
Where Kimi K2 Instruct pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+1.6points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+1.5points ahead
Every test, side by side
All 23 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2 Instruct
- SciCodeGLM-4.5 by 4.134.830.7+4.1
- SWE-bench VerifiedKimi K2 Instruct by 1.664.265.8+1.6
- Terminal-Bench HardKimi K2 Instruct by 1.52223.5+1.5
AgentsEven
- GDPValKimi K2 Instruct by 18.2018.2+18.2
- BrowseCompGLM-4.5 by 12.326.414.1+12.3
ReasoningGLM-4.5
- Humanity's Last ExamGLM-4.5 by 6.6136.4+6.6
- GPQA DiamondGLM-4.5 by 1.578.276.7+1.5
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationtie29.930.3tie
- AA-Omniscience · Accuracytie25.125.4tie
Other results11 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.
- AIME 2025GLM-4.5 by 43.893.349.5+43.8
- τ²-Bench Telecom (AA run)Kimi K2 Instruct by 30.44373.4+30.4
- AA Agentic IndexKimi K2 Instruct by 21.516.237.7+21.5
- LiveCodeBenchGLM-4.5 by 19.272.953.7+19.2
- Terminal-BenchGLM-4.5 by 7.537.530+7.5
- MMLU-ProGLM-4.5 by 3.584.681.1+3.5
- AA IntelligenceKimi K2 Instruct by 2.512.815.3+2.5
- SWE-Dev (Tools-allowed)GLM-4.5 by 1.363.261.9+1.3
- AA-Omnisciencetie-27.4-26.6tie
- MATH-500 (EM)tie98.297.4tie
- Artificial Analysis Coding Indextie26.325.9tie
Questions people ask
Which is better, GLM-4.5 or Kimi K2 Instruct?
GLM-4.5 wins two of the six areas where both have results: reasoning and following instructions. Kimi K2 Instruct wins coding. They are level on agents, facts and long documents.
Which is better for coding?
Kimi K2 Instruct. It wins 2 of the 3 coding tests both models report; GLM-4.5 wins 1.
Which is cheaper?
GLM-4.5 costs $0.60 per million input tokens and $2.20 per million output tokens; Kimi K2 Instruct costs $0.60 and $2.50. That makes GLM-4.5 about 10% cheaper for the same work.
How do you compare the two?
We use the 23 benchmark tests both models have published scores on. The verdict counts the 12 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 11 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.