Z AI

•

GLM-5.3

•

Open weights model

•

Released August 2026

GLM-5.3 (Max) API Provider Benchmarking & Analysis

Model Comparison

This analysis is intended to support you in choosing the best API provider of GLM-5.3 (Max) for your use-case.

Fastest

LithosAI (FP4 ULTRA CHAT)

752.9 t/s

Inco (FAST)

697.3 t/s

Supernova (FP4, FAST)

663.8 t/s

LithosAI (FP4)

535.2 t/s

Baseten (FAST)

306.5 t/s

Output speed

Total 26 providers

Lowest Latency

LithosAI (FP4 ULTRA CHAT)

3.27 s

Inco (FAST)

3.67 s

LithosAI (FP4)

4.33 s

Supernova (FP4, FAST)

4.48 s

Baseten (FAST)

7.07 s

Time to first answer token

Total 26 providers

Lowest Price

Supernova (FP4)

$0.41

DeepInfra

$0.45

Makora

$0.54

GMI

$0.63

LithosAI (FP4)

$0.68

Blended price (per 1M tokens)

Total 26 providers

GLM-5.3 (max) is available through 26 API providers, each offering different performance characteristics and pricing. Below is a comparison of the key metrics across providers.

For output speed, the top providers are LithosAI (FP4 ULTRA CHAT) (752.9 t/s), Inco (FAST) (697.3 t/s), and Supernova (FP4, FAST) (663.8 t/s). Speed varies significantly across providers, with a 1227% difference between the fastest and slowest.
For latency, LithosAI (FP4 ULTRA CHAT) (3.27s), Inco (FAST) (3.67s), and LithosAI (FP4) (4.33s) offer the lowest time to first answer token.
For pricing, Supernova (FP4) (0.41), DeepInfra (0.45), and Makora (0.54) offer the lowest blended prices per 1M tokens. Prices vary up to 6.2x across providers.
LithosAI (FP4 ULTRA CHAT) offers the best performance with both the highest speed and lowest latency. For cost optimization, Supernova (FP4) provides the most competitive pricing.

Speed

Output tokens per second · Higher is better

Latency

Seconds to first answer token received · Lower is better

Price

USD per 1M tokens (blended) · Lower is better

Update: Default performance benchmarking workload has updated to 10k input tokens to better reflect production use cases. You can still select different workloads above.

Endpoint Accuracy

Endpoint Accuracy Index: GLM-5.3 (Max)

v1.0 · Composite of BFCL v4-500, HLE-250 and AA-LCR-25 run against each provider endpoint · Percentage of the reference endpoint, with 95% confidence interval · Higher is better

Composite measure of how much of a model's accuracy a given provider endpoint preserves, from re-running BFCL v4-500, HLE-250 and AA-LCR-25 against that endpoint. Where a self-hosted reference endpoint exists, scores are expressed as a percentage of that reference (100 = matches reference); lower scores indicate accuracy lost to quantisation, sampling defaults, or other endpoint-side configuration. Scores are point-in-time snapshots. Methodology.

Pricing

Pricing: Cache Hit, Input, and Output: GLM-5.3 (max) Providers

Price (USD per M Tokens) · Lower is better · 10,000 input tokens

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

Cost per Intelligence Index Task: GLM-5.3 (max) Providers

Weighted average cost (USD) per Intelligence Index task · Segmented by token type · Lower is better

Weighted average cost per Intelligence Index task. Each evaluation’s cost is calculated from input, cache hit, cache write, reasoning, and answer token prices, divided by task count, and weighted by its Intelligence Index weight.

Output Speed vs. Price: GLM-5.3 (max) Providers

Blended at 7:2:1 (cache-input-output) · Output speed: output tokens per second · USD per 1M tokens (blended)

Most attractive quadrant

Pareto line

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Speed

Measured by Output Speed (tokens per second)

Output Speed: GLM-5.3 (Max)

Output speed: output tokens per second · 10,000 input tokens

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Figures represent performance of the model's first-party API or the median across providers where a first-party API is not available.

Time per Intelligence Index Task: GLM-5.3 (max) Providers

Weighted average decode time (minutes) per Intelligence Index task · Excludes TTFT and overhead time · Lower is better

The weighted average time (seconds) per Artificial Analysis Intelligence Index task. This is calculated by dividing output tokens per task by output speed, weighted by the relative weights of each benchmark in the Intelligence Index.

Cost per Task vs. Time per Task: GLM-5.3 (max) Providers

Weighted average cost (USD) per Intelligence Index task · Weighted average decode time (minutes) per Intelligence Index task

Most attractive quadrant

Pareto line

Latency

Measured by Time (seconds) to First Token

Time to First Answer Token: GLM-5.3 (max) Providers

Seconds to first token received · Lower is better · 10,000 input tokens

Time to first answer token received, in seconds, after API request sent. For reasoning models, this includes the 'thinking' time of the model before providing an answer. For models which do not support streaming, this represents time to receive the completion.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Cache Behaviour

Cache Hit Rate: GLM-5.3 (max) Providers

Share of cacheable input tokens served from cache · Median of the last four weeks, updated Sep 26, 2026

The share of cacheable input tokens served from the provider's prompt cache, measured during multi-turn agentic coding sessions. We regularly measure each endpoint and report a representative rate based on recent measurements. Higher values mean more of the input is billed at the cheaper cache-hit price, where the provider offers one, which lowers the effective cost of agentic workloads. For further details, see our methodology page.

Cost per Task vs. Cache Hit Rate: GLM-5.3 (max) Providers

Weighted average cost (USD) per Intelligence Index task · Cache hit rate: median of the last four weeks, updated Sep 26, 2026

Most attractive quadrant

Pareto line

End-to-End Response Time

Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed

End-to-End Response Time: GLM-5.3 (max) Providers

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · 10,000 input tokens

Seconds to receive a 500 token response. Key components:

Input time: Time to receive the first response token
Thinking time (only for reasoning models): Time reasoning models spend outputting tokens to reason prior to providing an answer. Amount of tokens based on the average reasoning tokens across a diverse set of 60 prompts (methodology details).
Answer time: Time to generate 500 output tokens, based on output speed

Key Comparison Metrics & API Features



Modal	100%	1M	Open	$1.97	108	0.84	23.97	18.50
Zai	99%	1M	Open	$2.01	77	2.74	35.30	26.05
DeepInfra	99%	1.05M	Open	$0.97	112	0.96	23.36	17.92
Modular	97%	1.05M	Open	$2.96	158	0.96	16.77	12.65
Parasail	98%	1M	Open	$1.97	152	1.04	17.47	13.15
Together AI	96%	1.05M	Open	$1.97	111	1.03	23.46	17.95
Inco	n/a	1M	Open	$4.08	670	0.83	4.56	2.99
Novita	96%	1M	Open	$2.00	96	1.62	27.53	20.73
Baseten	96%	1.05M	Open	$1.43	96	2.59	28.62	20.82
DigitalOcean	99%	1.05M	Open	$1.97	50	0.83	50.97	40.11
Makora	99%	1.05M	Open	$0.92	187	0.87	14.26	10.71
Databricks	96%	1M	Open	$2.73	202	0.84	13.23	9.92
Wafer	95%	1M	Open	$2.04	155	0.82	16.90	12.87
Baseten	94%	1M	Open	$3.75	293	0.56	9.08	6.82
LithosAI	97%	1.05M	Open	$2.35	412	0.65	6.72	4.86
LithosAI	n/a	1.05M	Open	$5.14	724	0.61	4.06	2.76
Inco	n/a	164k	Open	$2.04	122	1.71	22.18	16.38
Bitdeer AI	99%	1M	Open	$4.99	124	1.29	21.41	16.10
Crusoe	97%	786k	Open	$3.36	191	0.78	13.84	10.45
Supernova	100%	1M	Open	$1.19	261	1.63	11.19	7.65
FriendliAI	n/a	1M	Open	$1.81	162	1.18	16.63	12.36
Supernova	100%	1M	Open	$5.29	588	1.46	5.72	3.40
Nebius	100%	1M	Open	$8.68	229	1.49	12.40	8.73
GMI	97%	1M	Open	$1.42	73	2.02	36.49	27.58
Multiverse Computing	99%	1M	Open	$4.70	212	1.14	12.91	9.42
Scaleway	n/a	262k	Open	$12.94	258	1.29	10.98	7.76

Frequently Asked Questions

Common questions about GLM-5.3 (Max) providers

GLM-5.3 (Max) is available through 26 API providers: Modal, Zai, DeepInfra, Modular, Parasail, Together AI, Inco (FAST), Novita, Baseten, DigitalOcean, Makora, Databricks, Wafer, Baseten (FAST), LithosAI (FP4), LithosAI (FP4 ULTRA CHAT), Inco, Bitdeer AI, Crusoe (FP4), Supernova (FP4), FriendliAI, Supernova (FP4, FAST), Nebius (FP4), GMI, Multiverse Computing, and Scaleway. Each provider offers different performance characteristics and pricing.

GLM-5.3 (Max) is currently available through 26 API providers that we benchmark and track.

The fastest providers for GLM-5.3 (Max) by output speed are LithosAI (FP4 ULTRA CHAT) (752.9 t/s), Inco (FAST) (697.3 t/s), and Supernova (FP4, FAST) (663.8 t/s). Output speed measures how quickly tokens are generated after the model starts responding.

The providers with the lowest time to first answer token for GLM-5.3 (Max) are LithosAI (FP4 ULTRA CHAT) (3.27s), Inco (FAST) (3.67s), and LithosAI (FP4) (4.33s). Lower latency means faster initial response time.

The most affordable providers for GLM-5.3 (Max) by blended price are Supernova (FP4) ($0.41 per 1M tokens), DeepInfra ($0.45 per 1M tokens), and Makora ($0.54 per 1M tokens). Blended price uses a 7:2:1 cache hit/input/output token ratio.

The providers with the lowest input token pricing for GLM-5.3 (Max) are Makora ($0.14 per 1M input tokens), DeepInfra ($0.56 per 1M input tokens), and Supernova (FP4) ($0.63 per 1M input tokens).

The providers with the lowest output token pricing for GLM-5.3 (Max) are Supernova (FP4) ($1.98 per 1M output tokens), DeepInfra ($2.50 per 1M output tokens), and GMI ($3.08 per 1M output tokens).

Prices for GLM-5.3 (Max) vary up to 6.2x across providers. The most affordable is Supernova (FP4) at $0.41 per 1M tokens, while Scaleway charges $2.52 per 1M tokens.

Output speed for GLM-5.3 (Max) varies significantly across providers. LithosAI (FP4 ULTRA CHAT) is the fastest at 752.9 t/s, which is 13.3x faster than DigitalOcean at 56.8 t/s.

All 26 providers of GLM-5.3 (Max) support JSON mode for structured output.

All 26 providers of GLM-5.3 (Max) support function calling (tool use).

For GLM-5.3 (Max), LithosAI (FP4 ULTRA CHAT) offers the best performance with highest speed and lowest latency. For cost optimization, Supernova (FP4) provides the most competitive pricing.

When choosing a provider for GLM-5.3 (Max), consider: output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and API features like JSON mode or function calling.

Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.

For information about GLM-5.3 (Max)'s intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview

GLM-5.3 (Max) API Provider Benchmarking & Analysis

Fastest

Lowest Latency

Lowest Price

Speed

Latency

Price

Endpoint Accuracy

Endpoint Accuracy Index: GLM-5.3 (Max)

Endpoint Accuracy Index

Pricing

Pricing: Cache Hit, Input, and Output: GLM-5.3 (max) Providers

Cache Hit

Cost per Intelligence Index Task: GLM-5.3 (max) Providers

Cost per Intelligence Index Task

Output Speed vs. Price: GLM-5.3 (max) Providers

Output Speed

Speed

Output Speed: GLM-5.3 (Max)

Output Speed

Model Performance Representation

Time per Intelligence Index Task: GLM-5.3 (max) Providers

Time per Intelligence Index Task

Cost per Task vs. Time per Task: GLM-5.3 (max) Providers

Cost per Intelligence Index Task

Latency

Time to First Answer Token: GLM-5.3 (max) Providers

Time to First Answer Token

Median

Cache Behaviour

Cache Hit Rate: GLM-5.3 (max) Providers

What Cache Hit Rate Shows

Cost per Task vs. Cache Hit Rate: GLM-5.3 (max) Providers

Cost per Intelligence Index Task

End-to-End Response Time

End-to-End Response Time: GLM-5.3 (max) Providers

End-to-End Response Time

Key Comparison Metrics & API Features

Frequently Asked Questions

Where can I access GLM-5.3 (Max)?

How many API providers offer GLM-5.3 (Max)?

Which provider is fastest for GLM-5.3 (Max)?

Which provider has the lowest latency for GLM-5.3 (Max)?

Which provider is cheapest for GLM-5.3 (Max)?

Which provider has the lowest input price for GLM-5.3 (Max)?

Which provider has the lowest output price for GLM-5.3 (Max)?

How much do prices vary across GLM-5.3 (Max) providers?

How much does speed vary across GLM-5.3 (Max) providers?

Which GLM-5.3 (Max) providers support JSON mode?

Which GLM-5.3 (Max) providers support function calling?

Which is the best provider for GLM-5.3 (Max)?

How do I choose a provider for GLM-5.3 (Max)?

Does provider performance for GLM-5.3 (Max) change over time?

What are the overall capabilities of GLM-5.3 (Max)?