GLM 4.5 Air — Benchmarks

Benchmark scores for GLM 4.5 Air aggregated from public leaderboards (5 boards), with how it ranks among open models and how it compares with its nearest neighbours. See hardware requirements for what you need to run it.

llmrun Composite Score →

96.3

90% range 85.5–107.1 · from 4 boards

CodingNot enough data yet
AgentsNot enough data yet
MathNot enough data yet
ScienceNot enough data yet
ReasoningNot enough data yet

GLM 4.5 Air reasoning benchmarks

Multi-step logic and problem-solving — can the model think through a hard problem rather than recall an answer.

ForecastBench
ForecastBench, from the Forecasting Research Institute, asks models to predict real-world future events, with questions that only resolve after the model's training cutoff so the answers can't have been memorised. It is scored as a difficulty-adjusted Brier index, where higher is better.
59.2
59.2
median 58.1open leader 61.9
Overall best: O3 (Apr 16, 2025, unspecified) 62.5
68th percentile · #9 of 26 open · #45 of 82 all
#9 of 26
#45 of 82
Lab-reported
Epoch AI · Jul 2025

GLM 4.5 Air coding benchmarks

Writing, editing, and fixing real code — measured by whether the resulting program actually runs and passes tests.

GSO-Bench
GSO-Bench, built by a team of UC Berkeley-affiliated researchers, gives a coding agent a real codebase and a performance test and asks it to optimize runtime efficiency to match an expert developer's speedup without breaking correctness, across 102 tasks in 10 codebases and five languages. The scored metric, Opt@1, is the share of tasks where a single attempt reaches at least 95% of the human speedup while still passing all tests.
2.9%
2.9%
median 4.9%open leader 4.9%
Overall best: Claude Fable 5.1 (unspecified) 88.2%
lowest · #3 of 3 open · #39 of 42 all
#3 of 3
#39 of 42
Lab-reported
Epoch AI · Jul 2025

GLM 4.5 Air knowledge benchmarks

Breadth and depth of factual knowledge across many subjects, paired with the reasoning to apply it.

Humanity's Last Exam
Humanity's Last Exam (HLE) is a set of extremely difficult, expert-written questions across many fields, designed so that even frontier models score low. It is built to stay hard as models improve, measuring the true knowledge frontier.
8.1%
8.1%
median 8.2%open leader 24.4%
Overall best: GPT 6 Astra (unspecified) 54.8%
33rd percentile · #3 of 4 open · #37 of 51 all
#3 of 4
#37 of 51
Lab-reported
Epoch AI · Jul 2025
MMLU-Pro
MMLU-Pro is a harder, cleaned-up successor to MMLU with ten answer choices and more reasoning-heavy questions across 14 subjects, measuring broad knowledge and reasoning together.
81.4%
81.4%
median 52.1%open leader 88.0%
Overall best: Gemini-3.1-Pro 91.2%
88th percentile · #19 of 145 open · #56 of 257 all
#19 of 145
#56 of 257
Independent
TIGER-Lab

GLM 4.5 Air instruction following benchmarks

How precisely a model obeys explicit, verifiable instructions — format, length, and content constraints — and, more broadly, how much people prefer its responses head-to-head in blind human voting.

LMArena Text
Arena (arena.ai, formerly LMArena / Chatbot Arena) ranks models by millions of blind, head-to-head human votes on real prompts. The Arena Score shown here is a Bradley-Terry rating computed with Arena's default style control, so higher means people preferred that model's answers more often — it's a relative rating, not a percentage, so compare it only to other models on this board.
1373.0
1373.0
median 1321.5open leader 1487.1
Overall best: Gemini 4 Argon (High) 1523.9
67th percentile · #63 of 187 open · #187 of 405 all
#63 of 187
#187 of 405
Arena
Arena · Oct 2026

Score vs VRAM in the ≤96 GB tier

GLM 4.5 Air
Other models
Best Score for its VRAM
x: VRAM (GB, log scale) · y: llmrun Score

#30 of 60 that fit ≤96 GB

Each dot is a model at the best quant that fits ≤96 GB. The line joins the models with the best Score for their VRAM; a dot below it is beaten by a model that needs no more memory.

Full ≤96 GB ranking →

GLM 4.5 Air vs its neighbours

Same boards for the nearest open models per llmrun Score. The highest score in each row is marked.

Board	GLM 4.5 Airthis model	Qwen3.8 27Bbest in classQ4_K_M · 18 GB	GPT OSS 120Bnearest by ScoreQ4_K_M · 71 GB	Mistral Small 4 119B 2603nearest by ScoreQ4_K_M · 82 GB	NVIDIA Nemotron 3 Super 120B A12B BF16nearest by ScoreQ4_K_M · 75 GB
llmrun Score	96.3	122.1	97.5	96.4	95.8
MMLU-Pro	81.4%	—	80.8%	—	—
LMArena Text	1373.0	1437.8	1351.8	—	1360.1
Compare head to head		Compare	Compare	Compare	Compare

Scores aggregated from public benchmark sources (Epoch AI: 3 boards, Arena: 1 board, TIGER-Lab: 1 board) — each linked from the benchmark pages. llmrun does not run these benchmarks. Percentiles and ranks are among open models in the llmrun catalogue; the median and leader marks refer to the same field.