How accurate is GLM 5.3?

GLM 5.3 answered 54 of 80 questions correctly (67.5%) in our 2026-10-27 run, ranking #28 of 78 models.

GLM 5.3 generated a median of 154 output tokens per second in our 2026-10-27 run, ranking #16 of 78. Median time to first token was 0.8 s.

What is the latency of GLM 5.3?

GLM 5.3 had a median time to first token of 0.8 s in our 2026-10-27 run, ranking #40 of 78 models.

How much does GLM 5.3 cost?

GLM 5.3 cost $0.00093 per question on average as billed in our 2026-10-27 run, reasoning tokens included.

What is GLM 5.3 API pricing?

GLM 5.3 is listed at $0.079 per million input tokens and $1.4 per million output tokens on OpenRouter.

What is the context window of GLM 5.3?

GLM 5.3 accepts up to 1,048,576 tokens of context, according to OpenRouter's model catalogue.

Is GLM 5.3 open source?

Yes. GLM 5.3 is an open-weights model: Z.ai publishes the weights, so you can run it yourself as well as through an API.

GLM 5.3 is made by Z.ai. We tested it through OpenRouter using the model ID z-ai/glm-5.3.

Where can I use GLM 5.3 via API?

Through OpenRouter, GLM 5.3 is served by 8 hosts: AkashML, Baidu, Decart, Fireworks, Inceptron, InferenceNet, PrimeIntellect, Wafer.

Z.aiOpen weightsReleased 18 Aug 2026

GLM 5.3 benchmarks, speed and API pricing

Reviewed by LLM Verifications team · Tested 25 Oct 2026

Compare

Available through OpenRouter on

Wafer$0.0699 in · $1.4 out
Novita$0.63 in · $1.98 out
Sail Research$0.5 in · $2 out
Reka$0.14 in · $2.1 out
SiliconFlow$0.7 in · $2.2 out
GMICloud$0.7 in · $2.2 out
DeepInfra$0.5625 in · $2.5 out
Tenstorrent$0.6 in · $2.5 out
Phala$0.84 in · $2.64 out
Decart$0.84 in · $2.69 out
DigitalOcean$0.91 in · $2.86 out
DekaLLM$0.25 in · $2.95 out
Fireworks$0.98 in · $3.08 out
PrimeIntellect$0.98 in · $3.08 out
Inceptron$1.2 in · $3.5 out
InferenceNet$0.08 in · $3.6 out
Alibaba$1.19 in · $3.74 out
Friendli$1.26 in · $3.96 out
Morph$0.98 in · $4 out
CoreWeave$1.2 in · $4 out
Relace$0.0703 in · $4.3843 out
Makora$0.14 in · $4.4 out
AkashML$0.19 in · $4.4 out
Mistral$1.4 in · $4.4 out
Baidu$1.4 in · $4.4 out
BaseTen$1.4 in · $4.4 out
Nebius$1.4 in · $4.4 out
Crusoe$1.4 in · $4.4 out
Venice$1.4 in · $4.4 out
Together$1.4 in · $4.4 out
Parasail$1.4 in · $4.4 out
Modal$1.4 in · $4.4 out
Cloudflare$1.4 in · $4.4 out
AtlasCloud$1.4 in · $4.4 out
Z.AI$1.4 in · $4.4 out
Amazon Bedrock$1.68 in · $5.28 out
ModelRun$2.1 in · $6.6 out

Intelligence67.5% correct#28 of 72

Speed154tok/s#16 of 72

Cost$0.00093per question#25 of 72List $0.079 in · $1.4 out per 1M

Verbosity523tokens per question#15 of 72

Summary

Answers 54 of 80 verification questions correctly (67.5%), placing #28 of 72 models tested on the current suite.
Generates 154 output tokens per second, faster than 79% of the models we measure.
Starts responding after 0.8 s (median time to first token), including thinking time.
Costs $0.00093 per question as billed, #14 cheapest of 29 models priced $0.15 to $1 per 1M tokens.

Specifications

Reasoning: Yes
Input: text
Output: text
Context window: 1M tokens
Weights: Open weights
Knowledge cutoff: —
OpenRouter ID: z-ai/glm-5.3
Last tested: 25 Oct 2026

Peer group: 31 models with a blended list price $0.15 to $1 per 1M tokens (3 input : 1 output). Charts below compare GLM 5.3 with all models; cost and speed charts focus on this peer group.

Intelligence

Verification score: the share of 80 questions from open datasets answered correctly. See the benchmark datasets.

Score vs. cost per question

GLM 5.3 circled; dashed line joins models no cheaper model beats

Score vs. end-to-end response time

GLM 5.3 circled · accurate and quick sits top left

Score vs. tokens per question

GLM 5.3 circled · output tokens, reasoning included

Verification score

All models tested on the current suite

61 more models · see the full leaderboard

Cost

Average cost of one question, split into input, reasoning and answer tokens at each model's list price. Reasoning tokens are billed as output.

Cost per question, by token type

Peer group: blended price $0.15 to $1

1Llama 4 Scout$0.00025
2GPT-6 Luna$0.00025
3Ministral 3 8B 2512$0.00027
4Llama 3.3 70B Instruct$0.00031
5Ministral 3 14B 2512$0.00031
6GLM 5.3 Flash$0.00033
7gpt-oss-120b$0.00038
8Inkling Small$0.00047
9Claude Haiku 5.5$0.00048
10MiMo-V2.6-Flash$0.00056
11Llama 4 Maverick$0.00057
12GLM 5.3$0.00076
13DeepSeek V4.1 Flash 1017$0.00084
14MiniMax M3.1 Flash$0.0009
15DeepSeek V4 Flash Vision Exp$0.00092
16Hy3$0.0013
17Gemma 4 31B$0.0014
18Nemotron 3.5 Super VL$0.0014
19Qwen3 Coder Next$0.0017
20Solar Pro 4$0.0018
21Mistral Small 4$0.0019
22MiMo-V2.6-Pro$0.002
23Trinity Large Thinking$0.0025
24MiniMax M3$0.0027
25LongCat 2.0$0.0032
26Qwen3.7 Plus$0.0033
27Gemini 3.5 Flash Lite$0.0034
28Qwen3.6 35B A3B$0.0034
29Qwen3.8 27B$0.0049

Input Reasoning Answer

Cost per question, as billed

All models · lower is better

1Granite 4.0 Micro$0.000051
2gpt-oss-20b$0.000066
3Granite 4.2 8B$0.000077
4Nova Micro 1.0$0.000087
5Mercury 2.5$0.00011
6Phi 4$0.00012
7gpt-oss-120b$0.00013
8GLM 5.3 Flash$0.00015
9Ministral 3 3B 2512$0.00019
10GPT-6 Luna$0.00025
25GLM 5.3$0.00093

61 more models · see the full leaderboard

Speed and latency

Median of 3 streamed runs of the same prompt, measured 25 Oct 2026.

Output speed

Peer group ($0.15 to $1) · higher is better

17 more models · see the full leaderboard

Time to first token

Peer group ($0.15 to $1) · lower is better

16 more models · see the full leaderboard

End-to-end response time

Peer group ($0.15 to $1) · lower is better

17 more models · see the full leaderboard

Latency vs. output speed

GLM 5.3 circled · quick to start and fast sits top left

Results by category

Questions answered correctly in each category

Science20/30

Engineering10/15

Medicine5/11

Economics7/9

Law4/6

Humanities & social science8/9

More from Z.ai

Z.ai#28GLM 5.3 Flash67.5% verification score83 tok/s0.5 s first token$0.00015 / question

All Z.ai models →

Compare GLM 5.3

Head to head with its closest rivals, other Z.ai models and the top models overall.

Hosts and prices

Every host serving this model through OpenRouter, from OpenRouter's catalogue

Host	Input / 1M	Output / 1M	Context	Max output	Quantization
Wafer	$0.0699	$1.4	1M	944k	fp4
Novita	$0.63	$1.98	1M	131k	fp8
Sail Research	$0.5	$2	1M	944k	fp8
Reka	$0.14	$2.1	1M	262k	—
SiliconFlow	$0.7	$2.2	1M	262k	fp8
GMICloud	$0.7	$2.2	1M	944k	fp8
DeepInfra	$0.5625	$2.5	1M	131k	fp4
Tenstorrent	$0.6	$2.5	1M	131k	fp8
Phala	$0.84	$2.64	1M	944k	—
Decart	$0.84	$2.69	1M	944k	fp4
DigitalOcean	$0.91	$2.86	1M	128k	—
DekaLLM	$0.25	$2.95	1M	944k	—
Fireworks	$0.98	$3.08	1M	944k	—
PrimeIntellect	$0.98	$3.08	1M	131k	—
Inceptron	$1.2	$3.5	1M	944k	fp4
InferenceNet	$0.08	$3.6	1M	944k	fp4
Alibaba	$1.19	$3.74	1M	131k	—
Friendli	$1.26	$3.96	1M	944k	—
Morph	$0.98	$4	1M	944k	fp8
CoreWeave	$1.2	$4	1M	944k	fp8
Relace	$0.0703	$4.3843	1M	131k	—
Makora	$0.14	$4.4	1M	128k	fp4
AkashML	$0.19	$4.4	1M	131k	fp8
Mistral	$1.4	$4.4	1M	131k	nvfp4
Baidu	$1.4	$4.4	1M	131k	fp8
BaseTen	$1.4	$4.4	1M	262k	fp4
Nebius	$1.4	$4.4	1M	922k	fp4
Crusoe	$1.4	$4.4	1M	944k	fp4
Venice	$1.4	$4.4	1M	131k	—
Together	$1.4	$4.4	1M	944k	—
Parasail	$1.4	$4.4	1M	944k	fp8
Modal	$1.4	$4.4	1M	944k	—
Cloudflare	$1.4	$4.4	1M	944k	—
AtlasCloud	$1.4	$4.4	1M	131k	fp8
Z.AI	$1.4	$4.4	1M	131k	fp8
Amazon Bedrock	$1.68	$5.28	1M	131k	—
ModelRun	$2.1	$6.6	1M	524k	fp4

Sources

Scores, speed and billed cost are our own measurements. Prices, context window and hosts come from OpenRouter's catalogue on the test date.

GLM 5.3 on OpenRouter: list prices and hosts
Model card on Hugging Face: weights and licence
SuperGPQA (paper): 50 of the questions, ODC-BY 1.0
MMLU-Pro (paper): 30 of the questions, MIT
OpenRouter generation records: where billed cost and native token counts come from
Our methodology and editorial standards

GLM 5.3: frequently asked questions

How accurate is GLM 5.3?: GLM 5.3 answered 54 of 80 questions correctly (67.5%) in our 2026-10-27 run, ranking #28 of 78 models.
How fast is GLM 5.3?: GLM 5.3 generated a median of 154 output tokens per second in our 2026-10-27 run, ranking #16 of 78. Median time to first token was 0.8 s.
What is the latency of GLM 5.3?: GLM 5.3 had a median time to first token of 0.8 s in our 2026-10-27 run, ranking #40 of 78 models.
How much does GLM 5.3 cost?: GLM 5.3 cost $0.00093 per question on average as billed in our 2026-10-27 run, reasoning tokens included.
What is GLM 5.3 API pricing?: GLM 5.3 is listed at $0.079 per million input tokens and $1.4 per million output tokens on OpenRouter.
What is the context window of GLM 5.3?: GLM 5.3 accepts up to 1,048,576 tokens of context, according to OpenRouter's model catalogue.
Is GLM 5.3 open source?: Yes. GLM 5.3 is an open-weights model: Z.ai publishes the weights, so you can run it yourself as well as through an API.
Who makes GLM 5.3?: GLM 5.3 is made by Z.ai. We tested it through OpenRouter using the model ID z-ai/glm-5.3.
Where can I use GLM 5.3 via API?: Through OpenRouter, GLM 5.3 is served by 8 hosts: AkashML, Baidu, Decart, Fireworks, Inceptron, InferenceNet, PrimeIntellect, Wafer.

Z.aiOpen weightsReleased 18 Aug 2026

GLM 5.3 benchmarks, speed and API pricing

Reviewed by LLM Verifications team · Tested 25 Oct 2026

Compare

Available through OpenRouter on

Wafer$0.0699 in · $1.4 out
Novita$0.63 in · $1.98 out
Sail Research$0.5 in · $2 out
Reka$0.14 in · $2.1 out
SiliconFlow$0.7 in · $2.2 out
GMICloud$0.7 in · $2.2 out
DeepInfra$0.5625 in · $2.5 out
Tenstorrent$0.6 in · $2.5 out
Phala$0.84 in · $2.64 out
Decart$0.84 in · $2.69 out
DigitalOcean$0.91 in · $2.86 out
DekaLLM$0.25 in · $2.95 out
Fireworks$0.98 in · $3.08 out
PrimeIntellect$0.98 in · $3.08 out
Inceptron$1.2 in · $3.5 out
InferenceNet$0.08 in · $3.6 out
Alibaba$1.19 in · $3.74 out
Friendli$1.26 in · $3.96 out
Morph$0.98 in · $4 out
CoreWeave$1.2 in · $4 out
Relace$0.0703 in · $4.3843 out
Makora$0.14 in · $4.4 out
AkashML$0.19 in · $4.4 out
Mistral$1.4 in · $4.4 out
Baidu$1.4 in · $4.4 out
BaseTen$1.4 in · $4.4 out
Nebius$1.4 in · $4.4 out
Crusoe$1.4 in · $4.4 out
Venice$1.4 in · $4.4 out
Together$1.4 in · $4.4 out
Parasail$1.4 in · $4.4 out
Modal$1.4 in · $4.4 out
Cloudflare$1.4 in · $4.4 out
AtlasCloud$1.4 in · $4.4 out
Z.AI$1.4 in · $4.4 out
Amazon Bedrock$1.68 in · $5.28 out
ModelRun$2.1 in · $6.6 out

Intelligence67.5% correct#28 of 72

Speed154tok/s#16 of 72

Cost$0.00093per question#25 of 72List $0.079 in · $1.4 out per 1M

Verbosity523tokens per question#15 of 72

Summary

Answers 54 of 80 verification questions correctly (67.5%), placing #28 of 72 models tested on the current suite.
Generates 154 output tokens per second, faster than 79% of the models we measure.
Starts responding after 0.8 s (median time to first token), including thinking time.
Costs $0.00093 per question as billed, #14 cheapest of 29 models priced $0.15 to $1 per 1M tokens.

Specifications

Reasoning: Yes
Input: text
Output: text
Context window: 1M tokens
Weights: Open weights
Knowledge cutoff: —
OpenRouter ID: z-ai/glm-5.3
Last tested: 25 Oct 2026

Peer group: 31 models with a blended list price $0.15 to $1 per 1M tokens (3 input : 1 output). Charts below compare GLM 5.3 with all models; cost and speed charts focus on this peer group.

Intelligence

Verification score: the share of 80 questions from open datasets answered correctly. See the benchmark datasets.

Score vs. cost per question

GLM 5.3 circled; dashed line joins models no cheaper model beats

Score vs. end-to-end response time

GLM 5.3 circled · accurate and quick sits top left

Score vs. tokens per question

GLM 5.3 circled · output tokens, reasoning included

Verification score

All models tested on the current suite

61 more models · see the full leaderboard

Cost

Average cost of one question, split into input, reasoning and answer tokens at each model's list price. Reasoning tokens are billed as output.

Cost per question, by token type

Peer group: blended price $0.15 to $1

1Llama 4 Scout$0.00025
2GPT-6 Luna$0.00025
3Ministral 3 8B 2512$0.00027
4Llama 3.3 70B Instruct$0.00031
5Ministral 3 14B 2512$0.00031
6GLM 5.3 Flash$0.00033
7gpt-oss-120b$0.00038
8Inkling Small$0.00047
9Claude Haiku 5.5$0.00048
10MiMo-V2.6-Flash$0.00056
11Llama 4 Maverick$0.00057
12GLM 5.3$0.00076
13DeepSeek V4.1 Flash 1017$0.00084
14MiniMax M3.1 Flash$0.0009
15DeepSeek V4 Flash Vision Exp$0.00092
16Hy3$0.0013
17Gemma 4 31B$0.0014
18Nemotron 3.5 Super VL$0.0014
19Qwen3 Coder Next$0.0017
20Solar Pro 4$0.0018
21Mistral Small 4$0.0019
22MiMo-V2.6-Pro$0.002
23Trinity Large Thinking$0.0025
24MiniMax M3$0.0027
25LongCat 2.0$0.0032
26Qwen3.7 Plus$0.0033
27Gemini 3.5 Flash Lite$0.0034
28Qwen3.6 35B A3B$0.0034
29Qwen3.8 27B$0.0049

Input Reasoning Answer

Cost per question, as billed

All models · lower is better

1Granite 4.0 Micro$0.000051
2gpt-oss-20b$0.000066
3Granite 4.2 8B$0.000077
4Nova Micro 1.0$0.000087
5Mercury 2.5$0.00011
6Phi 4$0.00012
7gpt-oss-120b$0.00013
8GLM 5.3 Flash$0.00015
9Ministral 3 3B 2512$0.00019
10GPT-6 Luna$0.00025
25GLM 5.3$0.00093

61 more models · see the full leaderboard

Speed and latency

Median of 3 streamed runs of the same prompt, measured 25 Oct 2026.

Output speed

Peer group ($0.15 to $1) · higher is better

17 more models · see the full leaderboard

Time to first token

Peer group ($0.15 to $1) · lower is better

16 more models · see the full leaderboard

End-to-end response time

Peer group ($0.15 to $1) · lower is better

17 more models · see the full leaderboard

Latency vs. output speed

GLM 5.3 circled · quick to start and fast sits top left

Results by category

Questions answered correctly in each category

Science20/30

Engineering10/15

Medicine5/11

Economics7/9

Law4/6

Humanities & social science8/9

More from Z.ai

Z.ai#28GLM 5.3 Flash67.5% verification score83 tok/s0.5 s first token$0.00015 / question

All Z.ai models →

Compare GLM 5.3

Head to head with its closest rivals, other Z.ai models and the top models overall.

Hosts and prices

Every host serving this model through OpenRouter, from OpenRouter's catalogue

Host	Input / 1M	Output / 1M	Context	Max output	Quantization
Wafer	$0.0699	$1.4	1M	944k	fp4
Novita	$0.63	$1.98	1M	131k	fp8
Sail Research	$0.5	$2	1M	944k	fp8
Reka	$0.14	$2.1	1M	262k	—
SiliconFlow	$0.7	$2.2	1M	262k	fp8
GMICloud	$0.7	$2.2	1M	944k	fp8
DeepInfra	$0.5625	$2.5	1M	131k	fp4
Tenstorrent	$0.6	$2.5	1M	131k	fp8
Phala	$0.84	$2.64	1M	944k	—
Decart	$0.84	$2.69	1M	944k	fp4
DigitalOcean	$0.91	$2.86	1M	128k	—
DekaLLM	$0.25	$2.95	1M	944k	—
Fireworks	$0.98	$3.08	1M	944k	—
PrimeIntellect	$0.98	$3.08	1M	131k	—
Inceptron	$1.2	$3.5	1M	944k	fp4
InferenceNet	$0.08	$3.6	1M	944k	fp4
Alibaba	$1.19	$3.74	1M	131k	—
Friendli	$1.26	$3.96	1M	944k	—
Morph	$0.98	$4	1M	944k	fp8
CoreWeave	$1.2	$4	1M	944k	fp8
Relace	$0.0703	$4.3843	1M	131k	—
Makora	$0.14	$4.4	1M	128k	fp4
AkashML	$0.19	$4.4	1M	131k	fp8
Mistral	$1.4	$4.4	1M	131k	nvfp4
Baidu	$1.4	$4.4	1M	131k	fp8
BaseTen	$1.4	$4.4	1M	262k	fp4
Nebius	$1.4	$4.4	1M	922k	fp4
Crusoe	$1.4	$4.4	1M	944k	fp4
Venice	$1.4	$4.4	1M	131k	—
Together	$1.4	$4.4	1M	944k	—
Parasail	$1.4	$4.4	1M	944k	fp8
Modal	$1.4	$4.4	1M	944k	—
Cloudflare	$1.4	$4.4	1M	944k	—
AtlasCloud	$1.4	$4.4	1M	131k	fp8
Z.AI	$1.4	$4.4	1M	131k	fp8
Amazon Bedrock	$1.68	$5.28	1M	131k	—
ModelRun	$2.1	$6.6	1M	524k	fp4

Sources

Scores, speed and billed cost are our own measurements. Prices, context window and hosts come from OpenRouter's catalogue on the test date.

GLM 5.3 on OpenRouter: list prices and hosts
Model card on Hugging Face: weights and licence
SuperGPQA (paper): 50 of the questions, ODC-BY 1.0
MMLU-Pro (paper): 30 of the questions, MIT
OpenRouter generation records: where billed cost and native token counts come from
Our methodology and editorial standards

GLM 5.3: frequently asked questions

How accurate is GLM 5.3?: GLM 5.3 answered 54 of 80 questions correctly (67.5%) in our 2026-10-27 run, ranking #28 of 78 models.
How fast is GLM 5.3?: GLM 5.3 generated a median of 154 output tokens per second in our 2026-10-27 run, ranking #16 of 78. Median time to first token was 0.8 s.
What is the latency of GLM 5.3?: GLM 5.3 had a median time to first token of 0.8 s in our 2026-10-27 run, ranking #40 of 78 models.
How much does GLM 5.3 cost?: GLM 5.3 cost $0.00093 per question on average as billed in our 2026-10-27 run, reasoning tokens included.
What is GLM 5.3 API pricing?: GLM 5.3 is listed at $0.079 per million input tokens and $1.4 per million output tokens on OpenRouter.
What is the context window of GLM 5.3?: GLM 5.3 accepts up to 1,048,576 tokens of context, according to OpenRouter's model catalogue.
Is GLM 5.3 open source?: Yes. GLM 5.3 is an open-weights model: Z.ai publishes the weights, so you can run it yourself as well as through an API.
Who makes GLM 5.3?: GLM 5.3 is made by Z.ai. We tested it through OpenRouter using the model ID z-ai/glm-5.3.
Where can I use GLM 5.3 via API?: Through OpenRouter, GLM 5.3 is served by 8 hosts: AkashML, Baidu, Decart, Fireworks, Inceptron, InferenceNet, PrimeIntellect, Wafer.