GLM-4.5-Air vs MiniMax-01.
model vs model · prices as of
GLM-4.5-Air and MiniMax-01 split the 5 rounds this table counts: the two list price rows, the cheapest price we can explain, the standard price and context. The weighting you pick decides the rest.
Z.aiGLM-4.5-Air
$0.80/M tokens
cheapest we can explain, at Submodel · $0.09 per 1,000 pages
131k context · tool calling · open weights
MiniMaxMiniMax-01
$1.54/M tokens
cheapest we can explain, at NanoGPT · $0.151 per 1,000 pages
1M context · reads images · open weights
List price in (lower wins)
$0.20/M
$0.20/M
List price out (lower wins)
$1.10/M
$1.10/M
Cheapest we can explain (lower wins)
$0.80/M at Submodel
$1.54/M at NanoGPT
Standard price (lower wins)
$1.70/M
$1.70/M
Cost per 1,000 pages (lower wins)
$0.09
$0.151
Context
131k tokens
1M tokens
Open weights
yes
yes
Tool calling
yes
no
Structured output
not stated
no
Reads images
no
yes
Reasoning
yes
no
Released
28 Jul 2025
15 Jan 2025
Providers selling it
23
4
Price we measure against
the maker's own
the maker's own
Trains on your prompts (no is better)
not stated
not stated
Providers selling both · 3
NanoGPT$1.16 / $1.54
Kilo Gateway$1.24 / $1.70
OpenRouter$1.24 / $1.70
GLM-4.5-Air then MiniMax-01, per million tokens
Go further
Add a third or fourth model, or switch the weighting, in the tool.
What moved in the last seven days
- GLM-4.5-Air
Nous Portal$0.10 / $0.68 $0.13 / $0.8525% dearer ·
Frequently asked
- Which is cheaper, GLM-4.5-Air or MiniMax-01?
- GLM-4.5-Air, at $0.80 per million tokens against $1.54 for MiniMax-01, three parts input to one part output, read on 16 Sept 2026.
- Where is each one cheapest?
- GLM-4.5-Air: $0.80 at Submodel. MiniMax-01: $1.54 at NanoGPT.
- Can I buy both from one provider?
- Yes. 3 providers sell both: NanoGPT, Kilo Gateway, OpenRouter. The cheapest of them for the two together is NanoGPT.
- Which holds more context?
- MiniMax-01, at 1M tokens against 131k.
Prices exclude tax and are per million tokens, read 16 Sept 2026 from each provider's own catalog. The standard price blends three parts input to one part output; a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.