GLM-5.2 vs GLM-5.3.

model vs model · prices as of

GLM-5.2 takes 1 of 5 rounds against GLM-5.3. It is also the cheaper of the two to run, at $1.95 per million tokens against $2.60.

Z.aiwins 1 rounds

GLM-5.2

$1.95/M tokens

cheapest we can explain, at Crof · $0.243 per 1,000 pages

1M context · tool calling · open weights

Crof
Z.ai

GLM-5.3

$2.60/M tokens

cheapest we can explain, at Crof · $0.324 per 1,000 pages

1M context · tool calling · open weights

Crof

List price in (lower wins)

$1.40/M

$1.40/M

List price out (lower wins)

$4.40/M

$4.40/M

Cheapest we can explain (lower wins)

$1.95/M at Crof

$2.60/M at Crof

Standard price (lower wins)

$8.60/M

$8.60/M

Cost per 1,000 pages (lower wins)

$0.243

$0.324

Context

1M tokens

1M tokens

Open weights

yes

yes

Tool calling

yes

yes

Structured output

yes

yes

Reads images

no

no

Reasoning

yes

yes

Released

13 Jun 2026

14 Aug 2026

Providers selling it

94

73

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 · 57

  • Crof$1.95 / $2.60
  • Novita$3.99 / $5.59
  • TrustedRouter$3.74 / $6.10
  • DigitalOcean Gradient$4.30 / $6.25
  • Merge Gateway$6.45 / $4.30
  • Nous Portal$5.53 / $5.60
  • DeepInfra$4.65 / $7.60
  • Requesty$4.65 / $7.60

GLM-5.2 then GLM-5.3, per million tokens · showing the 8 cheapest of 57

Go further

Add a third or fourth model, or switch the weighting, in the tool.

What moved in the last seven days

  1. GLM-5.2Baidu$0.49 / $1.53 $1.40 / $4.40187% dearer ·
  2. GLM-5.2Inceptron$1.08 / $2.89 $1.10 / $2.992% dearer ·
  3. GLM-5.2Inceptron$1.08 / $2.89 $1.10 / $2.992% dearer ·
  4. GLM-5.2io.net Intelligence$1.68 / $5.28 $1.58 / $4.976% cheaper ·
  5. GLM-5.2Nous Portal$0.72 / $2.26 $0.90 / $2.8325% dearer ·
  6. GLM-5.2Novita$0.68 / $2.15 $0.65 / $2.045% cheaper ·
  7. GLM-5.2OpenRouter$0.97 / $3.04 $1.40 / $4.4045% dearer ·
  8. GLM-5.2Requesty$1.40 / $4.40 $0.75 / $2.4046% cheaper ·
  9. GLM-5.2StreamLake$0.48 / $1.51 $0.64 / $2.0234% dearer ·
  10. GLM-5.3AkashML$1.17 / $3.96 $0.97 / $3.3017% cheaper ·
  11. GLM-5.3DigitalOcean Gradient$1.40 / $4.40 $0.95 / $3.4027% cheaper ·
  12. GLM-5.3GMI Cloud$1.12 / $3.52 $1.05 / $3.306% cheaper ·
  13. GLM-5.3Inceptron$0.87 / $3.36 $1.07 / $4.0922% dearer ·
  14. GLM-5.3Inceptron$0.87 / $3.36 $1.07 / $4.0922% dearer ·
  15. GLM-5.3io.net Intelligence$1.50 / $4.75 $1.36 / $4.319% cheaper ·
  16. GLM-5.3Morph$0.97 / $3.31 $1.19 / $3.7418% dearer ·
  17. GLM-5.3Nous Portal$0.56 / $1.76 $0.88 / $2.9763% dearer ·
  18. GLM-5.3Novita$1.09 / $3.43 $0.91 / $2.8617% cheaper ·
  19. GLM-5.3Reka$1.01 / $3.41 $0.88 / $2.9713% cheaper ·
  20. GLM-5.3Requesty$1.20 / $4.20 $1.20 / $43% cheaper ·
  21. GLM-5.3Sail Research$1.40 / $4.40 $1.13 / $3.5619% cheaper ·
  22. GLM-5.3SiliconFlow$1.40 / $4.40 $1.12 / $3.5220% cheaper ·
  23. GLM-5.3TrustedRouter$1.03 / $3.25 $0.95 / $3.254% cheaper ·
  24. GLM-5.3Venice$1.75 / $5.50 $1.40 / $4.4020% cheaper ·
  25. GLM-5.3Wafer$1.19 / $4.40 $0.95 / $4.409% cheaper ·
  26. GLM-5.2Zenifra$7.20 / $22.50 $5.70 / $2016% cheaper ·
  27. GLM-5.3Eden AI$1.10 / $3.50 $1.40 / $4.4026% dearer ·

Frequently asked

Which is cheaper, GLM-5.2 or GLM-5.3?
GLM-5.2, at $1.95 per million tokens against $2.60 for GLM-5.3, three parts input to one part output, read on 16 Sept 2026.
Where is each one cheapest?
GLM-5.2: $1.95 at Crof. GLM-5.3: $2.60 at Crof.
Can I buy both from one provider?
Yes. 57 providers sell both: Crof, Novita, TrustedRouter, DigitalOcean Gradient, Merge Gateway, Nous Portal, and more. The cheapest of them for the two together is Crof.
Which holds more context?
Both hold 1M tokens in one call.

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.

Related head-to-headsGLM-5.1 vs GLM-5.2GLM-5.1 vs GLM-5.3GLM-5.2 vs GPT-5.6 Terra ProGLM-5.2 vs QVQ MaxMorecompare any two modelsevery model we track