GLM-5.2 vs Qwen3.8 Max Preview
Wins 0 of 7 areas
—
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
Qwen3.8 Max Preview is the stronger all-rounder.GLM-5.2 is cheaper.
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 software14Qwen3.8 Max Preview4 of 6 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own14Qwen3.8 Max Preview4 of 5 tests
- MathCompetition and research-level math03Qwen3.8 Max Preview3 of 3 tests
- ReasoningHard problems that need careful thinking13Qwen3.8 Max Preview3 of 4 tests
- Following instructionsDoing exactly what it is asked02Qwen3.8 Max Preview2 of 2 tests
- FactsGetting facts right instead of making them up12Qwen3.8 Max Preview2 of 3 tests
- Long documentsFinding answers in very long texts01Qwen3.8 Max Preview1 of 1 test
Long documents rests on a single test.
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-5.2 costs 27% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Qwen3.8 Max Preview pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+37.9points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+17points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+15.5points ahead
Where GLM-5.2 pulls ahead
- Programming problems, refreshed regularlyLiveBench · Coding+6.8points ahead
- Research-level physics problemsCritPt+3.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+2.5points 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.
CodingQwen3.8 Max Preview
- LiveBench · Agentic CodingQwen3.8 Max Preview by 12.951.864.7+12.9
- Terminal-Bench 2.1Qwen3.8 Max Preview by 10.977.988.8+10.9
- LiveBench · CodingGLM-5.2 by 6.879.772.9+6.8
- LMArena · WebDevQwen3.8 Max Preview by 68 rating points16041672+68 rating
- SWE-bench ProQwen3.8 Max Preview by 5.662.167.7+5.6
- SciCodetie51.252.1tie
AgentsQwen3.8 Max Preview
- Terminal-Bench 4.0Qwen3.8 Max Preview by 37.9138.9+37.9
- GDPValQwen3.8 Max Preview by 14.743.658.3+14.7
- τ-Bench V3 · BankingQwen3.8 Max Preview by 13.234.647.8+13.2
- AA ApexAgentsQwen3.8 Max Preview by 8.733.742.4+8.7
- AA IT-Bench SREGLM-5.2 by 2.442.740.3+2.4
ReasoningQwen3.8 Max Preview
- LiveBench · ReasoningQwen3.8 Max Preview by 9.678.688.2+9.6
- GPQA DiamondQwen3.8 Max Preview by 3.389.592.8+3.3
- CritPtGLM-5.2 by 3.220.917.7+3.2
- Humanity's Last ExamQwen3.8 Max Preview by 241.143.1+2
FactsQwen3.8 Max Preview
- SimpleQA VerifiedQwen3.8 Max Preview by 13.134.247.3+13.1
- AA-Omniscience · AccuracyQwen3.8 Max Preview by 7.424.331.7+7.4
- AA-Omniscience · Non-hallucinationGLM-5.2 by 2.573.771.2+2.5
MathQwen3.8 Max Preview
- FrontierMath Tier 4Qwen3.8 Max Preview by 1729.346.3+17
- FrontierMath Tiers 1-3 (v2)Qwen3.8 Max Preview by 15.559.274.7+15.5
- LiveBench · MathematicsQwen3.8 Max Preview by 1.589.891.3+1.5
Long documentsQwen3.8 Max Preview
- AA-LCRQwen3.8 Max Preview by 278.380.3+2
Following instructionsQwen3.8 Max Preview
- LiveBench · Instruction FollowingQwen3.8 Max Preview by 11.862.374.1+11.8
- IFBenchQwen3.8 Max Preview by 9.573.382.8+9.5
Other results19 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.
- Agents' Last ExamQwen3.8 Max Preview by 28.623.852.4+28.6
- ToolathlonQwen3.8 Max Preview by 24.348.272.5+24.3
- GDPval-AA v2Qwen3.8 Max Preview by 231 rating points15081739+231 rating
- DeepSWE 1.1Qwen3.8 Max Preview by 12.64456.6+12.6
- Toolathlon VerifiedQwen3.8 Max Preview by 12.659.972.5+12.6
- AA IntelligenceQwen3.8 Max Preview by 11.733.745.4+11.7
- DeepSWEQwen3.8 Max Preview by 10.446.256.6+10.4
- AA Agentic IndexQwen3.8 Max Preview by 10.239.449.6+10.2
- JobBenchQwen3.8 Max Preview by 1043.453.4+10
- AA-OmniscienceQwen3.8 Max Preview by 7.64.412+7.6
- NL2RepoQwen3.8 Max Preview by 748.955.9+7
- livebench_data_analysisQwen3.8 Max Preview by 4.773.778.4+4.7
- livebench_languageQwen3.8 Max Preview by 3.576.279.7+3.5
- Agents' Last Exam (ALE-CLI)Qwen3.8 Max Preview by 3.223.827+3.2
- Artificial Analysis Coding IndexQwen3.8 Max Preview by 368.871.8+3
- HLE (with tools)Qwen3.8 Max Preview by 1.554.756.2+1.5
- CyberGymQwen3.8 Max Preview by 1.377.278.5+1.3
- ProgramBench (Almost Solved)Qwen3.8 Max Preview by 19.510.5+1
- MLS Bench Litelower is bettertie40.441tie
Questions people ask
Which is better, GLM-5.2 or Qwen3.8 Max Preview?
Qwen3.8 Max Preview wins all seven areas where both have results: coding, agents, reasoning, facts, math, long documents and following instructions. GLM-5.2 wins none, but costs 27% less.
Which is better for coding?
Qwen3.8 Max Preview. It wins 4 of the 6 coding tests both models report; GLM-5.2 wins 1, and 1 is a tie.
Which is cheaper?
GLM-5.2 costs $1.40 per million input tokens and $4.40 per million output tokens; Qwen3.8 Max Preview costs $2.00 and $6.00. That makes GLM-5.2 about 27% 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 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 19 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.