Skip to main content

Z.ai: GLM 5.1

Z.ai: GLM 5.1 is z-ai's model. On the un-leaked meo benchmark it ranks #18 of 22 with 36.0% accuracy and an Effective Value (𝕍) of 0.0000 — earning a correct answer for about $0.1058 in 434s each.

as of 2026-06-08

meo rank
#18
of 22
Accuracy
36.0%
overall
𝕍 index (top=100)
0.003
raw 𝕍 0.0000 · N=10
$ / correct
$0.1058
434s each

Per-domain accuracy

Theory of mind
100%
Regex automaton
100%
Tape machine
53%
CSP (unsatisfiable)
50%
Watson-Glaser
50%
Framework bias
33%
Base-rate bias
10%
Logic / math / CS
3%
State tracking
0%
Long arithmetic
0%

Efficiency

Cost per correct
$0.1058
Seconds per correct
434
Tokens per correct
26,241
Correct answers
43 / 151
Total benchmark cost
$4.55
Avg tokens / item
7,473

Aggregated third-party benchmarks

Artificial Analysis (artificialanalysis.ai). Redistribution requires an AA commercial license.

Model details

Lab
z-ai
Released
2026-04-07
Weights
Open
Context
203K
Max output
Knowledge cutoff
Price in
$0.98/M
Price out
$3.08/M
Modality
text->text

Sources: openrouter · artificial_analysis · meo

FAQ

How accurate is Z.ai: GLM 5.1?

Z.ai: GLM 5.1 scores 36.0% overall on the meo benchmark (rank #18 of 22), an un-leaked private holdout graded objective-first.

What is Z.ai: GLM 5.1's Effective Value?

Its Effective Value (𝕍) is 0.0000 at N=10, ω=1, δ=1.5 — a fused metric of accuracy, speed, cost, and error-cascade. 𝕍 rankings shift with task depth.

How much does Z.ai: GLM 5.1 cost per correct answer?

About $0.1058 per correct answer (26,241 tokens, 434 seconds each) at benchmark-time pricing via OpenRouter.

Is Z.ai: GLM 5.1 open weights?

Yes — Z.ai: GLM 5.1 is released with open weights.