GLM-5.2 vs GPT-6.1 Sol
Wins 0 of 7 areas
—
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
GPT-6.1 Sol is the stronger all-rounder.GLM-5.2 is cheaper.
Scores updated · 27 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.
- ReasoningHard problems that need careful thinking05GPT-6.1 Sol5 of 5 tests
- CodingWriting and fixing software03GPT-6.1 Sol3 of 4 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own03GPT-6.1 Sol3 of 3 tests
- MathCompetition and research-level math03GPT-6.1 Sol3 of 3 tests
- FactsGetting facts right instead of making them up12GPT-6.1 Sol2 of 3 tests
- Long documentsFinding answers in very long texts01GPT-6.1 Sol1 of 1 test
- Following instructionsDoing exactly what it is asked01GPT-6.1 Sol1 of 1 test
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-5.2 costs 52% 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 GPT-6.1 Sol pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+71.4points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+70.7points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+55.1points ahead
Where GLM-5.2 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+28points ahead
Every test, side by side
All 27 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-6.1 Sol
- LMArena · WebDevGPT-6.1 Sol by 145 rating points16061751+145 rating
- SciCodeGPT-6.1 Sol by 351.254.2+3
- LiveBench · Agentic CodingGPT-6.1 Sol by 2.851.854.5+2.8
- LiveBench · Codingtie79.780.4tie
AgentsGPT-6.1 Sol
- Terminal-Bench 4.0GPT-6.1 Sol by 55.1156.1+55.1
- AA ApexAgentsGPT-6.1 Sol by 26.333.760+26.3
- GDPValGPT-6.1 Sol by 11.143.654.7+11.1
ReasoningGPT-6.1 Sol
- ARC-AGI-2GPT-6.1 Sol by 71.422.894.2+71.4
- SimpleBenchGPT-6.1 Sol by 24.158.882.9+24.1
- LiveBench · ReasoningGPT-6.1 Sol by 1478.692.6+14
- Humanity's Last ExamGPT-6.1 Sol by 11.841.152.9+11.8
- CritPtGPT-6.1 Sol by 10.820.931.7+10.8
FactsGPT-6.1 Sol
- SimpleQA VerifiedGPT-6.1 Sol by 39.734.273.9+39.7
- AA-Omniscience · AccuracyGPT-6.1 Sol by 37.824.362.1+37.8
- AA-Omniscience · Non-hallucinationGLM-5.2 by 2873.745.7+28
MathGPT-6.1 Sol
- FrontierMath Tier 4GPT-6.1 Sol by 70.729.3100+70.7
- FrontierMath Tiers 1-3 (v2)GPT-6.1 Sol by 34.559.293.7+34.5
- LiveBench · MathematicsGPT-6.1 Sol by 789.896.8+7
Following instructionsGPT-6.1 Sol
- LiveBench · Instruction FollowingGPT-6.1 Sol by 11.962.374.2+11.9
Other results7 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.
- AA-OmniscienceGPT-6.1 Sol by 37.14.441.5+37.1
- DeepSWE 1.1GPT-6.1 Sol by 31.24475.2+31.2
- ARC-AGI-1GPT-6.1 Sol by 19.57796.5+19.5
- HiL-Bench (Tools-allowed)GPT-6.1 Sol by 18.643.762.3+18.6
- AA IntelligenceGPT-6.1 Sol by 18.133.751.8+18.1
- livebench_languageGPT-6.1 Sol by 13.976.290.1+13.9
- livebench_data_analysisGPT-6.1 Sol by 973.782.7+9
Questions people ask
Which is better, GLM-5.2 or GPT-6.1 Sol?
GPT-6.1 Sol 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 52% less.
Which is better for coding?
GPT-6.1 Sol. It wins 3 of the 4 coding tests both models report; GLM-5.2 wins none, 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; GPT-6.1 Sol costs $2.00 and $10.00. That makes GLM-5.2 about 52% cheaper for the same work.
How do you compare the two?
We use the 27 benchmark tests both models have published scores on. The verdict counts the 20 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 7 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.