GLM 5.3 API

Zhipu's GLM 5.3, through an OpenAI-compatible API. 11 providers quote it now, and each request goes to the offer with the lowest estimated charge that meets your rules.

Providers compete for each request. The exchange ranks the offers that meet your rules by estimated charge. It fixes your maximum charge before the first token, and you pay for the tokens that the model uses.
The OpenAI and Anthropic APIs. Use the OpenAI or Anthropic SDK that you have. Set the base URL and your key, and send glm-5.3 as the model.
Your data rules. Send x-liquid-retention: none and only offers whose provider does not keep your content compete for the request.

Facts

Model name: glm-5.3
Maker: Zhipu
Context window: 200K – 1M tokens
Input price: $0.120 to $2.050 per million tokens
Output price: $0.035 to $6.200 per million tokens
Providers: 11 independent providers
Regions: asia, cn, eu, global, us, us-east, us-west

Providers

blackfuel: global, $0.120 input and $0.041 output per million tokens, $0.024 cache read per million tokens, 1M tokens of context
engy: asia, $0.150 input and $0.035 output per million tokens, $0.050 cache read per million tokens, 294.9K tokens of context
akash: us-west, $0.190 input and $4.400 output per million tokens, $0.190 cache read per million tokens, 1M tokens of context
Alpha.sh: us-west, $0.700 input and $2.300 output per million tokens, $0.700 cache read per million tokens, 1M tokens of context
Valar: us-east, $0.800 input and $2.500 output per million tokens, $0.100 cache read per million tokens, 1M tokens of context
deeprelay: cn, $0.900 input and $2.830 output per million tokens, $0.167 cache read per million tokens, 1M tokens of context
MarsCompute: global, $1.400 input and $4.400 output per million tokens, $1.400 cache read per million tokens, 1M tokens of context
EmpirioLabs AI: global, $1.400 input and $4.400 output per million tokens, $1.400 cache read per million tokens, 1M tokens of context
blackfuel: eu, $1.715 input and $4.410 output per million tokens, $0.343 cache read per million tokens, 1M tokens of context
TensorX: eu, $1.750 input and $4.500 output per million tokens, $0.438 cache read per million tokens, 1M tokens of context
SOKKAN: eu, $2.050 input and $6.200 output per million tokens, $2.050 cache read per million tokens, 200K tokens of context
Pera: us, $0.285 input and $2.825 output per million tokens, $0.285 cache read per million tokens, $0.02 fee for each request, 1M tokens of context

Scores

Intelligence: 31.7, #16 of 41
HLE: 42.3, #34 of 273
GPQA: 91.7, #29 of 250
AA-LCR: 79.7, #47 of 237
SciCode: 59.0, #11 of 123
SimpleQA Verified: 41.0, #52 of 126

Scores: Artificial Analysis (https://artificialanalysis.ai/)

Scores: Epoch AI, 'Capabilities & Benchmarking' (https://epoch.ai/benchmarks)

Request

curl https://router.inference.ai.exchange/v1/chat/completions \
  -H "Authorization: Bearer $LIQUID_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3",
    "messages": [{"role":"user","content":"Say hello."}]
  }'

Questions

What is GLM 5.3?: Subconscious is an inference provider designed for long-horizon agents. For requests that require 200k+ tokens, we're faster, more accurate, and cheaper than other providers thanks system level optimization and novel caching enhancements.
How much does GLM 5.3 cost?: Offers now ask $0.120 to $2.050 per million input tokens and $0.035 to $6.200 per million output tokens. A job pays for the tokens it uses. The exchange picks the cheapest eligible offer.
What is the context length of GLM 5.3?: The largest offer takes 1,048,576 input tokens. One job can generate up to 1,048,576 tokens. A smaller offer can take less. The exchange skips an offer that cannot take your message.
What can GLM 5.3 do?: Its providers list Reasoning budget, Reasoning effort, JSON object, JSON schema, Frequency penalty, Logit bias, Log probabilities, Min-p sampling, Presence penalty, Repetition penalty, Seed, Stop sequences, Top-k sampling, Top-p sampling, Text, Tools.
What inputs and outputs does GLM 5.3 support?: Text → Text.
Which providers serve GLM 5.3?: 11 independent providers quote it now, with 12 offers. The table above ranks them by the cost of a typical job.
How do I call GLM 5.3?: Send "glm-5.3" as the model in an OpenAI or Anthropic request. The API section links to setup steps for each client.
When was GLM 5.3 listed?: The exchange first listed it on 10/7/2026. This is not its release date.