GLM-5.3 vs Kimi K3
Wins 1 of 7 areas
Agents
Wins 6 of 7 areas
Coding · Reasoning · Facts · Math · Long documents · Following instructions
Kimi K3 is the stronger all-rounder.GLM-5.3 is cheaper and better at agents.
Both rank among the ten best models we track in agents.
Scores updated · 43 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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.
- CodingWriting and fixing software04Kimi K34 of 5 tests · 1 tie
- ReasoningHard problems that need careful thinking14Kimi K34 of 5 tests
- FactsGetting facts right instead of making them up12Kimi K32 of 3 tests
- MathCompetition and research-level math12Kimi K32 of 3 tests
- Long documentsFinding answers in very long texts01Kimi K31 of 1 test
- Following instructionsDoing exactly what it is asked01Kimi K31 of 1 test
- AgentsCarrying out multi-step tasks on its own31GLM-5.33 of 4 tests
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-5.3 costs 68% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Kimi K3 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+13.7points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+9.7points ahead
- Short factual questions, answered correctlySimpleQA Verified+9.6points ahead
Where GLM-5.3 pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+29.3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+23.6points ahead
- Real work tasks from 44 professionsGDPVal+5.8points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K3
- LMArena · WebDevKimi K3 by 33 rating points16221655+33 rating
- LiveBench · CodingKimi K3 by 2.57981.5+2.5
- LiveBench · Agentic CodingKimi K3 by 1.360.962.2+1.3
- Terminal-Bench 2.1Kimi K3 by 1.183.985+1.1
- SciCodetie5959.5tie
AgentsGLM-5.3
- Terminal-Bench 4.0GLM-5.3 by 29.341.912.6+29.3
- GDPValGLM-5.3 by 5.857.551.7+5.8
- τ-Bench V3 · BankingGLM-5.3 by 4.350.346+4.3
- AA IT-Bench SREKimi K3 by 1.646.147.7+1.6
ReasoningKimi K3
- SimpleBenchGLM-5.3 by 5.566.260.7+5.5
- LiveBench · ReasoningKimi K3 by 4.985.890.7+4.9
- Humanity's Last ExamKimi K3 by 4.642.346.9+4.6
- CritPtKimi K3 by 4.319.123.4+4.3
- GPQA DiamondKimi K3 by 1.891.793.5+1.8
FactsKimi K3
- AA-Omniscience · Non-hallucinationGLM-5.3 by 23.670.446.8+23.6
- AA-Omniscience · AccuracyKimi K3 by 13.733.947.6+13.7
- SimpleQA VerifiedKimi K3 by 9.64150.6+9.6
MathKimi K3
- FrontierMath Tier 4Kimi K3 by 9.729.339+9.7
- LiveBench · MathematicsGLM-5.3 by 3.587.984.4+3.5
- FrontierMath Tiers 1-3 (v2)Kimi K3 by 3.468.872.2+3.4
Following instructionsKimi K3
- LiveBench · Instruction FollowingKimi K3 by 2.169.371.4+2.1
Other results21 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.
- Program BenchKimi K3 by 58.81977.8+58.8
- Terminal-bench 3.0GLM-5.3 by 10.928.317.4+10.9
- GDPval-AA v2GLM-5.3 by 87 rating points17691682+87 rating
- livebench_data_analysisKimi K3 by 8.570.278.7+8.5
- livebench_languageKimi K3 by 5.679.985.5+5.6
- AA-OmniscienceKimi K3 by 5.414.319.7+5.4
- CyberGymGLM-5.3 by 4.584.580+4.5
- Toolathlon VerifiedKimi K3 by 3.57376.5+3.5
- PostTrainBenchGLM-5.3 by 3.239.836.6+3.2
- AA Agentic IndexGLM-5.3 by 2.853.450.6+2.8
- HLE (with tools)GLM-5.3 by 2.762.559.8+2.7
- DeepSWE 1.1Kimi K3 by 2.166.969+2.1
- ProgramBench (Almost Solved)GLM-5.3 by 1.51917.5+1.5
- Artificial Analysis Coding IndexKimi K3 by 1.474.876.2+1.4
- AA IntelligenceGLM-5.3 by 1.244.843.6+1.2
- Agents' Last Exam (ALE-CLI)tie28.527.6tie
- DeepSWEtie66.967.5tie
- SWE-Marathontie42.542tie
- Agents' Last Examtie28.528.3tie
- Toolathlontie7373.2tie
- NL2Repotie5858tie
Questions people ask
Which is better, GLM-5.3 or Kimi K3?
Kimi K3 wins six of the seven areas where both have results: coding, reasoning, facts, math, long documents and following instructions. GLM-5.3 wins agents, and costs 68% less.
Which is better for coding?
Kimi K3. It wins 4 of the 5 coding tests both models report; GLM-5.3 wins none, and 1 is a tie.
Which is cheaper?
GLM-5.3 costs $1.40 per million input tokens and $4.40 per million output tokens; Kimi K3 costs $3.00 and $15.00. That makes GLM-5.3 about 68% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 21 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.