Z.ai · in Borg

GLM-4.5-Air

GLM-4.5-Air is the leaner member of the first GLM-4.5 agent-oriented pair.

Reasoning131,072 tokenstext

131,072 tokensContext window

—Median time to first token (gateway proxy)

12.1Output tokens per second (whole turn)

$0.2 / $1.1Vendor API price in / out · per 1M

Included in Borg

Download Borg Borg for Business

At a glance

Specification	Value
Provider	Z.ai
Family	GLM
Release date	2025-07-28 Vendor source
Context window	131,072 tokens Vendor source
Input modalities	text
Output modalities	text
Reasoning	Yes
Open weights	Yes
Borg lane ids	`glm-4.5-air`, `z-ai/glm-4.5-air`
Median tokens per second (end to end) on Borg	12.1 output tokens per second

Not published by the vendor: thinking modes, knowledge cutoff, fusion roles.

What GLM-4.5-Air is good for

GLM-4.5-Air is the leaner member of the first GLM-4.5 agent-oriented pair. Z.ai presents it as the cost-effective alternative to the larger GLM-4.5: it uses fewer active parameters while retaining the family's long text context, tool use and switchable thinking and non-thinking modes. It makes sense for repeated code edits, lightweight agent steps, retrieval-assisted answers and prototype assistants where throughput or provider-side cost matters more than maximum reasoning headroom. It is a public model, not just a gateway spelling: Z.ai names it in the GLM-4.5 documentation and in its own pricing table. Borg includes it with the subscription; Fusion can route work to this smaller option, while direct lane selection keeps a budget-oriented workflow on the same model.

cost-conscious coding-agent steps
high-volume tool-call workflows
long-document text extraction
fast non-thinking draft responses

How Borg uses it

GLM-4.5-Air is available as a subscription catalog lane in Borg. The catalog does not publish which Fusion profiles route to this model, so no routing role is claimed here. You can pick it directly by setting the OpenAI-compatible gateway model field to a listed lane id. Check the current catalog for availability before building a workflow around a lane.

model: "glm-4.5-air"

Measured on Borg

Metric	Median	95th percentile	Samples
Time to first token (gateway proxy)	—	—	1
Tokens per second (end to end)	12.1 output tokens per second	22.3 output tokens per second	149

Output tokens per second (whole turn)

Higher is fasterFleet median 31.9 tokens/s

GLM-4.5-Air12.1 tokens/s

Time to first token (gateway proxy): time until the first streamed response event reaches the gateway; our closest available measure of time to first token
Tokens per second (end to end): tokens per second, end to end: provider-reported output tokens divided by the whole turn's elapsed time, including queueing and reasoning; not a streaming per-token rate

Completion speed covers the whole turn including queueing and reasoning time; it is not a streaming tokens-per-second figure.

Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 157 requests. Minimum 100 samples per published request metric.

Pricing

Included in Borg subscription. Vendor API list price: $0.2 input / $1.1 output per 1M tokens. Vendor price source.

Compare subscription and API pricing · Cost Savings Calculator

FAQ

What is GLM-4.5-Air?

GLM-4.5-Air is listed in Borg's subscription catalog as glm-4.5-air. GLM-4.5-Air is the leaner member of the first GLM-4.5 agent-oriented pair.

Is GLM-4.5-Air in Borg?

Yes. GLM-4.5-Air is listed in Borg's subscription catalog under the lane glm-4.5-air. You can choose the lane directly, or explore the available Fusion profiles. Availability and model details can change; check the current catalog before deployment.

What context window does GLM-4.5-Air have?

The published context window is 131,072 tokens. Consult the cited sources for details and current limits. Vendor source

Is GLM-4.5-Air good for coding?

No vendor-published coding evaluation is listed here. Its listed uses include cost-conscious coding-agent steps, high-volume tool-call workflows, long-document text extraction, fast non-thinking draft responses. Try the lane for your own tasks, and compare the output with your requirements before choosing a default.

How fast is GLM-4.5-Air on Borg?