ModelsCompareBest forBenchmarksStatusPricingAPI
Compare / head-to-head

GLM-5.2vsGLM-5.3-Flash

GLM-5.3-Flash leads 9 of 12 shared benchmarks. GLM-5.3-Flash is 8.9x cheaper per token. Both offer a 1M-token context window.

Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.

Shared benchmarks
3 – 9
GLM-5.3-Flash leads
Cheaper per token
GLM-5.3-Flash
8.9x cheaper, input + output
Larger context
Tie
both 1M tokens
Provider uptime (30d)
Z.ai
GLM-5.2
Z.ai · released 2026-06-24
text
Context
1M
Max out
131K
Input /1M
$1.4
Output /1M
$4.4
Cached /1M
$0.26
Scores
26 · 9 core
GLM-5.3-Flash
Z.ai · released 2026-09-16
textvision
Context
1M
Max out
131K
Input /1M
$0.15
Output /1M
$0.5
Cached /1M
$0.03
Scores
19 · 6 core
Quality

Benchmark matrix

BenchmarkGLM-5.2GLM-5.3-FlashΔEdge
Reported by both · 12
GPQA Diamond91.9% epoch90.2% epoch+1.7 ptGLM-5.2
Agents' Last Exam (ALE-CLI)23.8% 26.3% -2.5 ptGLM-5.3-Flash
AutomationBench v1.0.626.2% 48.8% -22.6 ptGLM-5.3-Flash
DeepSWE v1.146.2% 63.4% -17.2 ptGLM-5.3-Flash
FrontierMath Tier 4 v2 (Epoch AI run)29.3% epoch17.1% epoch+12.2 ptGLM-5.2
FrontierMath Tiers 1-3 v2 (Epoch AI run)59.2% epoch55.8% epoch+3.4 ptGLM-5.2
GDPval-AA v21508 1773 -265GLM-5.3-Flash
HLE (with tools)54.7% 55.3% -0.6 ptGLM-5.3-Flash
NL2Repo48.9% 56.3% -7.4 ptGLM-5.3-Flash
OTIS Mock AIME 2024-2025 (Epoch AI run)86.4% epoch93.9% epoch-7.5 ptGLM-5.3-Flash
Terminal-Bench 2.181% 84.3% -3.3 ptGLM-5.3-Flash
Toolathlon Verified59.9% 78.4% -18.5 ptGLM-5.3-Flash
Only GLM-5.2 reports · 13
AIME 202690% matharenanot reported
CyberGym77.2% not reported
ExploitBench24.4% not reported
ExploitGym 2h29 not reported
ExploitGym 6h39 not reported
FrontierSWE67.5% not reported
PostTrainBench31.7% not reported
ProgramBench (Almost Solved)9.5% not reported
SimpleQA Verified34.2% epochnot reported
SWE-Bench Pro62.1% not reported
SWE-bench Verified (Epoch AI run)78.7% epochnot reported
SWE-Marathon v1.119.4% not reported
Terminal-Bench 3.04.6% not reported
Only GLM-5.3-Flash reports · 7
LMArena Elonot reported1471.9
BabyVisionnot reported53.4%
Chartography (with tools)not reported78%
CharXiv Reasoning (with tools)not reported89.4%
MMVUnot reported80.5%
MVBenchnot reported77.8%
OfficeQA Pronot reported62.4%
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GLM-5.2 minus GLM-5.3-Flash in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing

Side by side

SpecGLM-5.2GLM-5.3-FlashEdge
Context window1M tokens1M tokensTie
Max output131K tokens131K tokensTie
Input price / 1M$1.4 $0.15 GLM-5.3-Flash
Output price / 1M$4.4 $0.5 GLM-5.3-Flash
Cached input / 1M$0.26 $0.03 GLM-5.3-Flash
Input + output / 1M
Lower is cheaper. List prices; batch, tool and regional fees excluded.
$5.8$0.65GLM-5.3-Flash
Modalitiestexttext · visionGLM-5.3-Flash
Released2026-06-242026-09-16
Cited benchmark scores2619
Reliability

Z.ai status

All providers
More matchups

GLM-5.2 vs …

More matchups

GLM-5.3-Flash vs …

Built by Respan
Which one wins on your data?

Public benchmarks are a starting point. Run GLM-5.2 and GLM-5.3-Flash on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.