GLM 5.3 Flash Pricing: $0.15 In, $0.50 Out

Short answer. GLM 5.3 Flash, a Z.ai model that Mistral serves through its API, costs $0.15 per million input tokens, $0.03 for cached input and $0.50 for output on Mistral’s pricing page on October 23, 2026. Z.ai’s own pricing page prints the same three prices. Mistral marks the model GA, dated October 22, 2026.

In plain words. A token is a short piece of text, about three-quarters of a word. GLM 5.3 Flash is a model made by Z.ai. Mistral and Z.ai both sell it, at the same price per token. A sample request costs about a ninth of what it costs on GLM 5.3.

GLM 5.3 Flash is the newest row in the “Generalist models” table of Mistral’s API pricing page, just above Z.ai GLM 5.3, whose Mistral page is dated September 15, 2026. Mistral’s model page describes it in three sentences: “Z.AI’s GLM 5.3 Flash is a multimodal 320B-parameter (18B active) Mixture-of-Experts model. Built for developer agent workflows, visual coding, and complex tool-use tasks. The model is served without Mistral modifications.” This post sets Mistral’s price next to Z.ai’s, explains a column of Z.ai’s table that is easy to misread, and works out what one request costs on GLM 5.3 Flash and on Mistral’s other low-priced models.

The price on Mistral and on Z.ai

Per million tokens	On Mistral	On Z.ai
Input	$0.15	$0.15
Cached input	$0.03	$0.03
Output	$0.50	$0.50
Cached input storage	no column	“Limited-time Free”

So the three prices match to the cent. That is not the case for every GLM model the two companies both sell. For GLM 5.3, Mistral’s row read “Z.ai GLM 5.3 ↗ | $1.4 | $0.14 | $4.4 |” and Z.ai’s read “GLM-5.3 | $1.4 | $0.26 | Limited-time Free | $4.4 |”: the same input and output prices, but a cached input price of $0.14 on Mistral and $0.26 on Z.ai. The cached input price is the price of input tokens the API has already seen in an earlier request and keeps in its cache.

“Limited-time Free” does not mean the model is free

Z.ai’s table has five columns, and “Limited-time Free” appears only in the fourth, “Cached Input Storage”. The other three price columns carry dollar figures: GLM 5.3 Flash costs $0.15 per million input tokens, $0.03 per million cached input tokens and $0.50 per million output tokens, whatever that fourth column says. The column’s name says what it covers: storing cached input. The page, as read on October 23, 2026, gives no other figure for it and no end date for the limited time.

Against GLM 5.3 and Mistral’s other low-priced models

All prices below are from the same table on Mistral’s pricing page, read on October 23, 2026, in US dollars per million tokens. The last column is a sample request with 10,000 input tokens and 1,000 output tokens, with no cache.

Model on Mistral	Input	Cached input	Output	10,000 in + 1,000 out
GLM 5.3 Flash	$0.15	$0.03	$0.50	$0.002
GLM 5.3	$1.40	$0.14	$4.40	$0.0184
Mistral Small 4	$0.15	$0.015	$0.60	$0.0021
Ministral 3 14B	$0.20	$0.02	$0.20	$0.0022
Ministral 3 8B	$0.15	$0.015	$0.15	$0.00165
Ministral 3 3B	$0.10	$0.01	$0.10	$0.0011

The page printed those rows as “Mistral Small 4 ↗ | $0.15 | $0.015 | $0.6 |”, “Ministral 3 14B ↗ | $0.2 | $0.02 | $0.2 |”, “Ministral 3 8B ↗ | $0.15 | $0.015 | $0.15 |” and “Ministral 3 3B ↗ | $0.1 | $0.01 | $0.1 |”.

The sample request, worked out for the two GLM models:

GLM 5.3 Flash: 10,000 × $0.15 ÷ 1,000,000 = $0.0015 for input, plus 1,000 × $0.50 ÷ 1,000,000 = $0.0005 for output, so $0.002.
GLM 5.3: 10,000 × $1.40 ÷ 1,000,000 = $0.014 for input, plus 1,000 × $4.40 ÷ 1,000,000 = $0.0044 for output, so $0.0184.

A thousand such requests cost $2.00 on GLM 5.3 Flash and $18.40 on GLM 5.3, which is 9.2 times as much. Price by price, GLM 5.3 costs 1.40 ÷ 0.15 = 9.33 times as much for input, 0.14 ÷ 0.03 = 4.67 times as much for cached input and 4.40 ÷ 0.50 = 8.8 times as much for output.

Against Mistral’s own small models the gaps are narrow. GLM 5.3 Flash has the same input price as Mistral Small 4 and Ministral 3 8B, and its output price of $0.50 sits between the Ministral 3 models ($0.10 to $0.20) and Mistral Small 4 ($0.60). Its cached input price, $0.03, is twice Mistral Small 4’s $0.015. On the sample request the six models span $0.0011 to $0.0184, and three of them, GLM 5.3 Flash, Mistral Small 4 and Ministral 3 14B, land within $0.0002 of each other.

The order depends on the shape of the request. Turn the sample around, to 1,000 input tokens and 10,000 output tokens, and GLM 5.3 Flash costs 1,000 × $0.15 ÷ 1,000,000 + 10,000 × $0.50 ÷ 1,000,000 = $0.00515, against $0.00615 on Mistral Small 4 and $0.0022 on Ministral 3 14B. These figures are prices only: they say nothing about how well each model does a given task.

What Mistral and Z.ai say about the model

Mistral’s model page, read on October 23, 2026, says “October 22, 2026 Blog License GA Third-party Open v 5.3-flash”, lists the identifier zai-glm-5-3-flash, and gives “Context 1M Max output 128k”. GA is General Availability, which Mistral’s model lifecycle page, read on October 22, 2026, describes as “Production-ready and stable.” Mistral Large 4, by contrast, is in Public Preview. “Third-party” means Mistral did not make the model, and the page adds that “The model is served without Mistral modifications.”

Z.ai’s page for the model gives the same limits and the kinds of data it takes and returns: “Input Modality Video / Image / Text / File Output Modality Text Context Length 1M Maximum Output Tokens 128K”. The output is text only, which is why the model belongs in the API pricing table, where every model is charged per token and returns text. Z.ai adds that “GLM-5.3-Flash has 320B total parameters with 18B activated.”

On Z.ai the model has siblings. Its page lists the model codes “glm-5.3-flash / glm-5.3-flashx / glm-5.3-flashx-legal”, and the pricing table has a row “GLM-5.3-FlashX | $0.37 | $0.075 | Limited-time Free | $1.25 |”. Mistral’s pricing page, on October 23, 2026, listed GLM 5.3 Flash only, not FlashX.

The GLM Coding Plan

Z.ai’s page for the model opens with one sentence about a subscription: “GLM-5.3-Flash is now fully available on the GLM Coding Plan . With native multimodal capabilities and 3× the quota, it delivers a smoother and more cost-effective coding experience.” This post quotes that sentence and nothing more about the plan. It compares API prices per token, and gives no plan price.

When we saw it

Mistral’s model page dates GLM 5.3 Flash October 22, 2026. This site saw it on Mistral’s pricing page on October 23, 2026, through its daily scan of that page, and added it the same day after reading both companies’ pages.

The model has its own page with every figure and its source: GLM 5.3 Flash. The other models in the comparison have theirs: GLM 5.3, Mistral Small 4, Ministral 3 14B, Ministral 3 8B and Ministral 3 3B. Every Mistral model, and those of the other providers, are in the API pricing table. Mistral’s other new model this month is covered in Mistral Large 4 API pricing.

How we checked

Mistral’s pricing page, the Mistral model pages of GLM 5.3 Flash and GLM 5.3, Z.ai’s pricing page and Z.ai’s GLM-5.3-Flash page were read by a script on October 23, 2026. Mistral’s page showed Standard prices in US dollars; Z.ai’s says its prices are in US dollars. GLM 5.3 Flash first appeared in the daily scan of Mistral’s pricing page on October 23, 2026; the scan only flags new models and never writes prices. Each price and line quoted here comes from the saved text of those readings, and a test checks every quote and every figure against it. The ratios and the cost of the sample requests are worked out from those prices, with the sums shown.

Frequently asked questions

How much does GLM 5.3 Flash cost on Mistral?

Mistral's pricing page, read on October 23, 2026, lists Z.ai GLM 5.3 Flash at $0.15 per million input tokens, $0.03 for cached input and $0.50 for output. Z.ai's own pricing page prints the same three prices.

Is GLM 5.3 Flash free on Z.ai?

No: Z.ai's pricing page charges $0.15 per million input tokens, $0.03 for cached input and $0.50 for output. The words 'Limited-time Free' sit in its Cached Input Storage column, not in the price of the model.

How does GLM 5.3 Flash compare with GLM 5.3 on Mistral?

On Mistral, GLM 5.3 costs $1.40 per million input tokens and $4.40 for output, 9.33 and 8.8 times the prices of GLM 5.3 Flash. A request with 10,000 input and 1,000 output tokens costs $0.002 on GLM 5.3 Flash and $0.0184 on GLM 5.3.

Drafted with AI assistance from the cited source and reviewed by Álvaro Lucero Rufino, who is responsible for what stays published here. Automated publishing was discontinued in September 2026; posts that no longer met that standard were removed rather than left up. This is original explanatory writing, not a reproduction — read the source above for the full story.

GLM 5.3 Flash API Pricing: $0.15 In, $0.50 Out, Same on Mistral and Z.ai