See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditGLM 4.5 Air is the lightweight member of Z.ai's GLM 4.5 family, a text-only model built for agent-centric applications where the work is tool calls and multi-step tasks rather than open-ended conversation.
01 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 126 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$0.13
field median $0.30
Output / 1M tokens
$0.85
field median $1.25
Context
131K
field median 524K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
GLM 4.5 Air is a compact text model aimed at agent workloads: calling tools, working through steps, producing structured output that another system consumes. It is the smaller sibling in the GLM 4.5 line, trading parameter count for cost and throughput. It does not handle images, audio or any other modality, so anything involving documents as pictures or screen understanding needs a different model.
Where it sits in its line, and when a sibling is the better pick.
GLM 4.5 Air first appeared in July 2025 alongside the wider GLM 4.5 release, positioned below the flagship as the cheaper option for the same agent-shaped jobs. Reach for it when you are running a high volume of similar agent steps and the per-call cost matters more than the last increment of reasoning quality. When a task is genuinely hard, single-shot and expensive to get wrong, the full GLM 4.5 is the better pick.
The API, the apps it powers, and what its limits let you do.
GLM 4.5 Air is reached through the Z.ai API and through the usual third-party model routers that carry open-weight families, so it slots into an existing agent framework without much rework. The context window is large enough to hold a long tool-call history, a set of retrieved documents and a detailed system prompt at the same time, and the generous output ceiling means it can write a long file or a full structured response in one pass rather than being stitched together across calls.
What changes because this exists, or why it does not.
A cheap model that holds a long agent trace changes what you are willing to run repeatedly. Loops that were uneconomic at flagship prices, checking every page of a site, classifying every ticket, retrying a plan three ways, become routine. Nothing here is novel on its own, the significance is that agent-shaped work moves into the price band where you stop counting calls.
Superseded by GLM 4.5V. See the whole line.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $0.13 | per 1M tokens | 2026-09-24 | Check |
| Output | $0.85 | per 1M tokens | 2026-09-24 | Check |
GLM 4.5 Air sits at the low end of the market, cheap enough that input cost effectively disappears from most budgets and output remains the only line worth watching. Against Western flagship models it is an order of magnitude less expensive, which is the whole argument for using it in high-volume agent loops.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $6.85 |
| A busy support assistant | 200M tokens | 40M tokens | $60.00 |
| A document pipeline | 1000M tokens | 100M tokens | $215.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 131,072 tokens |
|---|---|
| Maximum output | 98,304 tokens |
| Modalities | text |
| Released | 25/07/2025 |
| Status | Current |
| Catalogue identifier | z-ai/glm-4.5-air |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from Z.ai
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Came before
The first we track in this line.
Came after
GLM 4.5V →Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.
10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
Cheap agent models mean more automated systems reading your site more often, and reading more of it in one pass given the context they can hold. Your documentation, pricing pages and comparison content are now being ingested by loops that run continuously rather than by a person asking one question. Write pages that survive being read out of order and quoted in fragments, because that is how an agent will use them.
Most buyers meet GLM 4.5 Air indirectly rather than by name, through the Z.ai API or a model router inside a product someone else built. It is a backend choice, so the encounter is usually a support agent, a research tool or an internal assistant that happens to be running on it.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.