The Known Good Updated 25 Jul 2026
Microsoft: Phi 4 vs Z.ai: GLM 5 ·Microsoft vs Z.ai

Microsoft: Phi 4 vs Z.ai: GLM 5 — head-to-head comparison

Microsoft: Phi 4 profile Z.ai: GLM 5 profile All comparisons
Microsoft: Phi 4 — Microsoft
Quality
#50 of 61
33based on 3 of 6 evaluations
Known Good Index
Speed
#145 of 316
55.4
Output tokens per second
Price
#40 of 340
$0.09
Blended per 1M tokens
In $0.07 · Out $0.14
Cache hit price
unranked
per 1M tokens
Context
#334 of 345
16k
tokens (~25 pages)
Z.ai: GLM 5 — Z.ai
Quality
#19 of 61
77based on 4 of 6 evaluations
Known Good Index
Speed
#214 of 316
37.6
Output tokens per second
Price
#217 of 340
$1.35
Blended per 1M tokens
In $0.95 · Out $2.55
Cache hit price
#138 of 206
$0.20
per 1M tokens (−79%)
Context
#189 of 345
205k
tokens (~300 pages)
Summary

On the Known Good Index, Microsoft: Phi 4 scores 33 and Z.ai: GLM 5 scores 77 — a 44-point lead for Z.ai: GLM 5. They share 2 published evaluations, with Microsoft: Phi 4 ahead on 0 and Z.ai: GLM 5 ahead on 2.

Blended pricing is $0.09 per 1M tokens for Microsoft: Phi 4 and $1.35 for Z.ai: GLM 5 — Microsoft: Phi 4 is about 15.4× cheaper.

Throughput is 55 tokens per second for Microsoft: Phi 4 and 38 for Z.ai: GLM 5, with time to first token at 0.2s and 2.5s. Microsoft: Phi 4 accepts 16k tokens of context against 205k.

Microsoft: Phi 4 is open weights under Open weights (unrestricted); Z.ai: GLM 5 is open weights under Open weights (unrestricted). 14 tracked hosts serve one or both, and 1 serve both.

Generated from tracked data on 26 Jul 2026 · how these comparisons are computed →

At a glance
MetricMicrosoft: Phi 4Z.ai: GLM 5
Known Good Index 33 based on 3 of 6 evaluations 77 based on 4 of 6 evaluations
Blended $/1M $0.09 $1.35
Output tok/s 55 38
Context window 16k 205k
Providers 1 14
Jump to section

Quality

Known Good Index and the capability indices, side by side

IndicesComponents
Index scores

Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.

MetricMicrosoft: Phi 4Z.ai: GLM 5
Known Good Index 33 77
Coding Index 67
Mathematics Index 39 80
Science Index 53 92
Reasoning Index 47
Agentic Index 67
Openness Index 60 60
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = Microsoft: Phi 4 · marker = Z.ai: GLM 5 · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond56 / 88
MATH / AIME14 / 80

Right-hand figures read Microsoft: Phi 4 / Z.ai: GLM 5. Microsoft: Phi 4 leads on 0 of 2, Z.ai: GLM 5 on 2.

The Known Good

Arena Elo

Human preference rating from blind pairwise votes

Arena ratings

Elo from blind pairwise votes, ingested from Arena. Higher is better.

MetricMicrosoft: Phi 4Z.ai: GLM 5
Text Arena 1,217 1,445
Text Arena (style controlled) 1,256 1,457
The Known Good

Price & Cost

Published API pricing by token type

Per token typeCache discount
Published pricing

USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.

MetricMicrosoft: Phi 4Z.ai: GLM 5
Input $/1M $0.07 $0.95
Output $/1M $0.14 $2.55
Blended $/1M $0.09 $1.35
Cache read $/1M $0.200
The Known Good
We publish blended price rather than cost-per-task. Cost-per-task requires token counts from an evaluation suite we do not run. Blended price is computed from published rates and is fully reproducible — see the pricing methodology.

Context Window

Maximum input accepted and maximum output emitted

Context and output limits

Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.

MetricMicrosoft: Phi 4Z.ai: GLM 5
Context window 16k 205k
Max output tokens 16k 131k
The Known Good

Speed & Latency

Median throughput and time to first token across hosting providers

Output speedTime to first token
Throughput and latency

Medians across all tracked providers, derived from OpenRouter endpoint telemetry.

MetricMicrosoft: Phi 4Z.ai: GLM 5
Output tokens/s 55 38
Time to first token 0.2s 2.5s
The Known Good

Providers

Which hosts serve each model, and at what price and speed

14 hosts · 1 serve both
Full provider benchmarks →
Provider Microsoft: Phi 4 blendedMicrosoft: Phi 4 tok/s Z.ai: GLM 5 blendedZ.ai: GLM 5 tok/s
DeepInfra $0.09 38 $0.97 29
Amazon Bedrock · first party $1.55 116
GMICloud $0.93 60
Phala $1.77 57
Chutes $1.35 49
Venice $1.55 40
Baidu · first party $1.08 39
Parasail $1.55 39
AtlasCloud $1.50 36
Novita $1.55 36
StreamLake $0.93 36
Z.AI · first party $1.55 33
SiliconFlow $1.35 30
DigitalOcean $1.16 8

Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.

The Known Good
Sources for this page
Pricing & providers — OpenRouter Metadata — models.dev Benchmarks — Epoch AI (CC-BY 4.0)Elo — Arena Full attribution →