GLM-5.3-Flash vs Qwen3.8 Max Preview
Wins 0 of 8 areas
—
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Following instructions
Qwen3.8 Max Preview is the stronger all-rounder.GLM-5.3-Flash is cheaper.
Scores updated · 38 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. 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 thinking04Qwen3.8 Max Preview4 of 4 tests
- MathCompetition and research-level math03Qwen3.8 Max Preview3 of 3 tests
- CodingWriting and fixing software13Qwen3.8 Max Preview3 of 5 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own13Qwen3.8 Max Preview3 of 5 tests · 1 tie
- FactsGetting facts right instead of making them up12Qwen3.8 Max Preview2 of 3 tests
- Images and chartsUnderstanding pictures, charts and video01Qwen3.8 Max Preview1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.8 Max Preview1 of 1 test
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
Images and charts, 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-Flash costs 92% 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
- Completes tasks by operating a computer desktopOSWorld-Verified+71.3points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+29.2points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+21.3points ahead
Where GLM-5.3-Flash pulls ahead
- Finds the root cause of IT system incidentsAA IT-Bench SRE+10.9points ahead
- Programming problems, refreshed regularlyLiveBench · Coding+6.1points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+1.2points ahead
Every test, side by side
All 38 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 7.956.864.7+7.9
- LiveBench · CodingGLM-5.3-Flash by 6.17972.9+6.1
- LMArena · WebDevQwen3.8 Max Preview by 61 rating points16091670+61 rating
- Terminal-Bench 2.1Qwen3.8 Max Preview by 4.584.388.8+4.5
- SciCodetie51.652.1tie
AgentsQwen3.8 Max Preview
- OSWorld-VerifiedQwen3.8 Max Preview by 71.314.886.1+71.3
- AA IT-Bench SREGLM-5.3-Flash by 10.951.240.3+10.9
- Terminal-Bench 4.0Qwen3.8 Max Preview by 6.132.838.9+6.1
- GDPValQwen3.8 Max Preview by 1.157.258.3+1.1
- τ-Bench V3 · Bankingtie47.247.8tie
ReasoningQwen3.8 Max Preview
- LiveBench · ReasoningQwen3.8 Max Preview by 10.677.688.2+10.6
- Humanity's Last ExamQwen3.8 Max Preview by 3.239.943.1+3.2
- CritPtQwen3.8 Max Preview by 2.315.417.7+2.3
- GPQA DiamondQwen3.8 Max Preview by 1.691.292.8+1.6
FactsQwen3.8 Max Preview
- SimpleQA VerifiedQwen3.8 Max Preview by 15.33247.3+15.3
- AA-Omniscience · AccuracyQwen3.8 Max Preview by 4.227.531.7+4.2
- AA-Omniscience · Non-hallucinationGLM-5.3-Flash by 1.272.471.2+1.2
Images and chartsQwen3.8 Max Preview
- LMArena · VisionQwen3.8 Max Preview by 17 rating points12951312+17 rating
MathQwen3.8 Max Preview
- FrontierMath Tier 4Qwen3.8 Max Preview by 29.217.146.3+29.2
- FrontierMath Tiers 1-3 (v2)Qwen3.8 Max Preview by 18.955.874.7+18.9
- LiveBench · MathematicsQwen3.8 Max Preview by 10.181.291.3+10.1
Following instructionsQwen3.8 Max Preview
- LiveBench · Instruction FollowingQwen3.8 Max Preview by 21.352.874.1+21.3
Other results15 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 26.126.352.4+26.1
- GDP (Surge AI)Qwen3.8 Max Preview by 7.415.422.8+7.4
- JobBenchGLM-5.3-Flash by 7.360.753.4+7.3
- Terminal-Bench-Science 0.1Qwen3.8 Max Preview by 7.14.811.9+7.1
- DeepSWE 1.1GLM-5.3-Flash by 6.863.456.6+6.8
- ToolathlonGLM-5.3-Flash by 5.978.472.5+5.9
- HLE-DiamondQwen3.8 Max Preview by 5.818.224+5.8
- CritPt v2 HardQwen3.8 Max Preview by 524.529.5+5
- AA-OmniscienceQwen3.8 Max Preview by 4.57.512+4.5
- AA IntelligenceQwen3.8 Max Preview by 3.641.845.4+3.6
- livebench_languageQwen3.8 Max Preview by 2.477.379.7+2.4
- livebench_data_analysisQwen3.8 Max Preview by 276.478.4+2
- AA Agentic IndexGLM-5.3-Flash by 1.651.249.6+1.6
- NL2Repotie56.355.9tie
- Artificial Analysis Coding Indextie71.571.8tie
Questions people ask
Which is better, GLM-5.3-Flash or Qwen3.8 Max Preview?
Qwen3.8 Max Preview wins seven of the eight areas we test: coding, agents, reasoning, facts, images and charts, math and following instructions. GLM-5.3-Flash wins none, but costs 92% less. They are level on long documents.
Which is better for coding?
Qwen3.8 Max Preview. It wins 3 of the 5 coding tests both models report; GLM-5.3-Flash wins 1, and 1 is a tie.
Which is cheaper?
GLM-5.3-Flash costs $0.15 per million input tokens and $0.50 per million output tokens; Qwen3.8 Max Preview costs $2.00 and $6.00. That makes GLM-5.3-Flash about 92% cheaper for the same work.
How do you compare the two?
We use the 38 benchmark tests both models have published scores on. The verdict counts the 23 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.