GLM 5.3 API pricing

Zhipu · Budget price tier · GLM-5.3 License · released 2026-08. Prices in USD per 1 million tokens. Data snapshot: 2026-11-21.

Outlier listing. The price we quote is one host's rate with input near free and output at or above the going rate, not what most hosts charge, so a cost built on it can be far from what you would pay. The typical host charges about $1.05 per 1M input tokens for this model.

A support ticket workload of 100K requests a month costs $80.48 on GLM 5.3; the next cheaper model in the Budget price tier for this workload, Mistral Small 3.1 24B, does it for $80.40. The price we quote for GLM 5.3 has changed 22 times since 2026-08-19, most recently on 2026-11-16 when the quoted host changed (-60.7%); it is now the 135th cheapest of 160 models in the Budget price tier. None of its 35 hosts undercuts the price we quote. It is rated a stable release but does not meet all four criteria of the FinOps Friendly badge.

Available on

AWS Bedrock regions: ap-northeast-1, ap-southeast-1, eu-central-1, eu-west-1, us-east-1, us-east-2, us-west-2

Market position

On blended price ($1.001 per 1M tokens at 30% input, 70% output) it is cheaper than 16% of models in the Budget price tier: 135th cheapest of 160. Across every price tier it is cheaper than 56% of priced text models in the catalogue: 135th cheapest of 308. Its Arena ELO of 1475 ranks 19th of the 157 models that carry one, tied with 2 others, above 87% of them. With an Arena score of 1475, GLM 5.3 clears OptimNow's Strong and Standard working quality thresholds: Strong (Arena 1469+), used for support ticket, knowledge Q&A and marketing content; Standard (Arena 1419+), used for meeting summary, invoice processing and call summary. Each threshold is a distance below the best ranked model, which scores 1529 on Arena scores as of 11 Nov 2026.

Licence, size and capabilities

GLM 5.3 is open weights, which is not the same as open source: the weights are published, but under GLM-5.3 License, a licence that restricts some uses. Read it before self-hosting commercially. Its published size is 743B parameters. Listed capabilities: Text, Reasoning, Agents. Filling its 1M context window once costs $0.0699 in input tokens at list price.

Pricing by host

A host is a company that runs the model and sells it by the token: the model maker, a cloud platform like Amazon Bedrock, Azure or Vertex, or an independent host. One model, several hosts, and often several prices. This model is sold by 35 different hosts. The price we quote elsewhere on this page is marked. We quote one host's rate so that models stay comparable; it is not a recommendation to buy there.

The price we quote is already the cheapest of the 35 hosts: Wafer at $0.0699 input / $1.40 output per 1M tokens. It serves a quantized fp4 build, which is not the same product as the full-precision model. The dearest, ModelRun, charges 5.2x the cheapest.

HostInput $/1MOutput $/1MPrecision
Wafer (price we quote)$0.0699$1.40fp4
Decart$0.56$1.81fp4
SiliconFlow$0.70$1.90fp8
Novita$0.63$1.98fp8
Doubleword$0.70$2.20nvfp4
Fireworks$0.70$2.20not declared
Cloudflare$0.70$2.20not declared
DeepInfra$0.5625$2.50fp4
Tenstorrent$0.60$2.50fp8
DigitalOcean$0.91$2.86not declared
CoreWeave$0.75$3.00nvfp4
Sail Research$0.75$3.00fp8
Featherless$0.98$3.08not declared
InferenceNet$0.084$3.60fp4
Cohere$0.50$3.50nvfp4
Inceptron$1.05$3.50fp4
Alibaba$1.19$3.74not declared
Relace$0.0703$4.3843fp4
AkashML$0.19$4.40fp4
Friendli$1.26$3.96not declared
Z.AI (model maker)$1.40$4.40fp8
Nebius$1.40$4.40fp4
SambaNova$1.40$4.40not declared
Mistral$1.40$4.40nvfp4
Baidu$1.40$4.40fp8
BaseTen$1.40$4.40fp4
Crusoe$1.40$4.40fp4
Venice$1.40$4.40not declared
Together$1.40$4.40not declared
Parasail$1.40$4.40fp4
Modal$1.40$4.40not declared
Makora$1.50$4.40fp4
Morph$1.066$4.974fp8
Amazon Bedrock$1.68$5.28not declared
ModelRun$2.10$6.60fp4

List prices as published on 2026-11-21, via OpenRouter. Negotiated rates, committed-use discounts and provisioned capacity are not shown. These hosts do not all serve the model at the same numeric precision, so the cheapest is not a like-for-like substitute for the dearest. These listings are promotional rates, below the host's standard rate: Novita's, 55% off; DeepInfra's, 38% off; Featherless's, 30% off; Friendli's, 10% off; SambaNova's, 50% off; and Venice's, 20% off.

Monthly cost at 100K requests, by business use case

From the after-discounts figure (prompt caching on the cached input share, batch API for async workloads, only where Zhipu publishes those rates), which assumes you implement both, to the list price. One figure means no discount is published.

Use caseTokens (in/out)Monthly (after discounts – list)
Support Ticket1,500 / 500$80
Knowledge Q&A2,000 / 800$126
Meeting Summary10,000 / 1,200$238
Marketing Content2,500 / 1,800$269
Coding Task3,000 / 2,000$301
Invoice Processing1,500 / 600$94
Call Summary2,000 / 700$112
Agent Workflow6,000 / 3,000$462

Price history

GLM 5.3 entered our daily price archive on 2026-08-19; tracking began on 2026-08-15. It listed at $1.40 input / $4.40 output per 1M tokens that day. The price we quote has changed 22 times since then, 9 of them a change of quoted host rather than a repricing. The latest change, on 2026-11-16, was -60.7% on the blended price (30% input, 70% output), to $0.0699 input / $1.40 output (quoted host changed, not a repricing). Net of all changes, the blended price is -71.4% against 2026-08-19.

Price chart for GLM 5.3: the blended price we quote went from $3.50 on 2026-08-19 to $1.001, across 22 changes.Price we quoteRange across hostsQuoted host changed$0$2.5$5$7.5$1019 Aug12 Sep5 Oct29 Oct21 Nov
Blended price per 1M tokens (30% input, 70% output). The grey band runs from the cheapest to the dearest host each day, quantized builds included, recorded from 30 Aug. Of the 80 days the range covers, the price we quote was the cheapest host's rate on 22 and the dearest host's rate on 17.
DateOld price (in / out)New price (in / out)Blended change
2026-11-16$0.084 / $3.60$0.0699 / $1.40-60.7% (quoted host changed, not a repricing)
2026-11-14$0.0699 / $1.40$0.084 / $3.60+154.3% (quoted host changed, not a repricing)
2026-11-05$0.08 / $3.60$0.0699 / $1.40-60.7% (quoted host changed, not a repricing)
2026-11-01$0.0699 / $1.40$0.08 / $3.60+154.2%
2026-10-26$0.079 / $1.40$0.0699 / $1.40-0.3%

17 earlier changes are not listed.

Alternatives to GLM 5.3

3 of the 4 comparable models below cost less than GLM 5.3: GLM 5.3 Flash undercuts it by 93%, for 3 Arena points less.

The nearest models in the Budget price tier on Arena ELO, each with its blended price and score difference against GLM 5.3.

Other models from Zhipu

The 8 nearest to GLM 5.3 in price, of 16 other Zhipu models in the catalogue. Prices per 1M tokens, input then output.

All 17 Zhipu models and prices

Common questions about GLM 5.3 pricing

How much does GLM 5.3 cost?
GLM 5.3 costs $0.0699 per 1M input tokens and $1.40 per 1M output tokens at Wafer, the host whose rate we quote. Wafer serves a quantized fp4 build, which is not the same product as the full-precision model. The model maker, Z.AI, lists it at $1.40 per 1M input tokens and $4.40 per 1M output tokens. Prices as listed on 2026-11-21.
Which host is cheapest for GLM 5.3?
The price we quote is already the cheapest of the 35 hosts: Wafer at $0.0699 input / $1.40 output per 1M tokens. It serves a quantized fp4 build, which is not the same product as the full-precision model. The dearest, ModelRun, charges 5.2x the cheapest. Host rates as listed on 2026-11-21.
Does GLM 5.3 offer batch or cached pricing?
Yes, for prompt caching only. Cache reads cost $0.0689 per 1M tokens, as listed on 2026-11-21. No batch pricing is published.
What would GLM 5.3 cost per month?
With the Support Ticket profile (1,500 input and 500 output tokens per request), 100,000 requests a month cost about $80 on GLM 5.3 at the list price quoted on 2026-11-21.
Can GLM 5.3 be self-hosted?
Yes, within the terms of its licence. GLM 5.3 is open weights, which is not the same as open source: the weights are published, but under GLM-5.3 License, a licence that restricts some uses. Read it before self-hosting commercially.
Has the price of GLM 5.3 changed?
GLM 5.3 entered our daily price archive on 2026-08-19; tracking began on 2026-08-15. It listed at $1.40 input / $4.40 output per 1M tokens that day. The price we quote has changed 22 times since then, 9 of them a change of quoted host rather than a repricing. The latest change, on 2026-11-16, was -60.7% on the blended price (30% input, 70% output), to $0.0699 input / $1.40 output (quoted host changed, not a repricing). Net of all changes, the blended price is -71.4% against 2026-08-19.

Its weights are published, which changes what you may run yourself: the cost gap between open weights and proprietary APIs.

These are list prices. Your real cost also depends on caching and batching, negotiated discounts, marketplace agreements, and the contracts you already have. OptimNow audits exactly that.

Build the business case · Get a FinOps review

LLM pricing · Models by maker · Cloud compute pricing · Compare models · Price barometer · FinOps guides · API and MCP server