DeepSeek-V3.1 vs GLM-4.5
Wins 4 of 6 areas
Coding · Agents · Reasoning · Long documents
Wins 1 of 6 areas
Following instructions
DeepSeek-V3.1 is the stronger all-rounder.GLM-4.5 is better at following instructions.
Scores updated · 22 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.
- CodingWriting and fixing software30DeepSeek-V3.13 of 3 tests
- ReasoningHard problems that need careful thinking20DeepSeek-V3.12 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own20DeepSeek-V3.12 of 2 tests
- Long documentsFinding answers in very long texts10DeepSeek-V3.11 of 1 test
- Following instructionsDoing exactly what it is asked01GLM-4.51 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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.
DeepSeek-V3.1 costs 20% 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 DeepSeek-V3.1 pulls ahead
- Code for real scientific research problemsSciCode+4.3points ahead
- Reasons across sets of long documentsAA-LCR+4points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+3.9points ahead
Where GLM-4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+12.4points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+2.6points ahead
Every test, side by side
All 22 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-V3.1
- SciCodeDeepSeek-V3.1 by 4.339.134.8+4.3
- Terminal-Bench HardDeepSeek-V3.1 by 32522+3
- SWE-bench VerifiedDeepSeek-V3.1 by 1.86664.2+1.8
AgentsDeepSeek-V3.1
- GDPValDeepSeek-V3.1 by 5.65.60+5.6
- BrowseCompDeepSeek-V3.1 by 3.63026.4+3.6
ReasoningDeepSeek-V3.1
- CritPtDeepSeek-V3.1 by 220+2
- Humanity's Last ExamDeepSeek-V3.1 by 1.314.313+1.3
- GPQA Diamondtie77.978.2tie
FactsEven
- AA-Omniscience · Non-hallucinationGLM-4.5 by 12.417.529.9+12.4
- AA-Omniscience · AccuracyDeepSeek-V3.1 by 3.92925.1+3.9
Other results10 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 43.549.893.3+43.5
- LiveCodeBenchGLM-4.5 by 16.556.472.9+16.5
- SWE-Dev (Tools-allowed)GLM-4.5 by 9.953.363.2+9.9
- Terminal-BenchGLM-4.5 by 6.231.337.5+6.2
- τ²-Bench Telecom (AA run)GLM-4.5 by 5.637.443+5.6
- Artificial Analysis Coding IndexDeepSeek-V3.1 by 3.429.726.3+3.4
- AA Agentic IndexDeepSeek-V3.1 by 2.718.916.2+2.7
- AA-OmniscienceGLM-4.5 by 2.2-29.6-27.4+2.2
- MMLU-Protie83.784.6tie
- AA Intelligencetie13.512.8tie
Questions people ask
Which is better, DeepSeek-V3.1 or GLM-4.5?
DeepSeek-V3.1 wins four of the six areas where both have results: coding, agents, reasoning and long documents. GLM-4.5 wins following instructions. They are level on facts.
Which is better for coding?
DeepSeek-V3.1. It wins 3 of the 3 coding tests both models report; GLM-4.5 wins none.
Which is cheaper?
DeepSeek-V3.1 costs $0.56 per million input tokens and $1.68 per million output tokens; GLM-4.5 costs $0.60 and $2.20. That makes DeepSeek-V3.1 about 20% cheaper for the same work.
How do you compare the two?
We use the 22 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 10 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.