zaiCertifiedReleased 06/16/2026

GLM 5.2

GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks.

View API reference ↓
Input / output priceFrom$0.19 / $1.80per 1M
24h uptime99.96%
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.inferencebureau.com/v1",
  apiKey: process.env.IB_API_KEY,
});

const res = await client.chat.completions.create({
  model: "zai/glm-5.2",
  messages: [{ role: "user", content: "Why is the sky blue?" }],
});

console.log(res.choices[0].message.content);

Providers

Dispatch routes requests across these providers and fails over between them without changing the model. Restrict or order providers with team Directives. Prices are gateway list prices before the platform fee.

ProviderContextMax outputLatencyThroughputInputOutputCacheCapabilitiesZDRNo trainingRegionsStatus
alibaba1M128K1.4 s87 tps$1.40$4.40Read $0.28Write —5 featuresYesYesGlobalCertified
Baseten1M128K1.5 s83 tps$1.40$4.40Read $0.14Write —5 featuresYesYesUSCertified
DeepInfra1M128K1.2 s75 tps$0.56$1.80Read $0.11Write —5 featuresNoYesGlobalCertified
digitalocean262K52K0.5 s97 tps$0.70$2.20Read $0.11Write —4 featuresYesYesGlobalCertified
friendli1M128K0.2 s172 tps$1.40$4.40Read $0.26Write —5 featuresNoNoGlobalCertified
gmicloud1.05M1.05M2.3 s100 tps$1.40$4.40Read $0.26Write —5 featuresNoNoGlobalListed
inceptron1M1M0.5 s235 tps$1.20$4.20Read $0.26Write —5 featuresNoYesGlobalCertified
nebius1.02M1.02M0.5 s204 tps$1.40$4.40Read $0.14Write —5 featuresYesYesGlobalCertified
novita1M128K2.0 s101 tps$0.65$2.04Read $0.12Write —5 featuresNoNoGlobalCertified
parasail262K128K0.3 s167 tps$1.40$4.40Read $0.26Write —5 featuresYesYesGlobalCertified
runware1M128K1.5 s132 tps$1.12$3.61Read $0.23Write —5 featuresYesYesGlobalCertified
streamlake1M128K1.2 s71 tps$0.80$2.52Read $0.15Write —5 featuresNoNoGlobalCertified
tencent1M128K1.4 s50 tps$1.40$4.40Read $0.26Write —5 featuresNoNoGlobalListed
Together AI256K128K0.2 s253 tps$1.40$4.40Read $0.26Write —5 featuresYesYesGlobalCertified
wafer1M128K0.3 s162 tps$0.19$8.00Read $0.18Write —5 featuresYesYesUSCertified
zai1M128K1.1 s72 tps$1.40$4.40Read $0.26Write —5 featuresNoNoGlobalCertified

Uptime

24 hours

Request success rate per provider.

alibaba100.00%
Baseten99.91%
DeepInfra99.91%
digitalocean99.97%
friendli99.93%
gmicloud100.00%
inceptron100.00%
nebius100.00%
novita100.00%
parasail100.00%
runware99.89%
streamlake99.93%
tencent99.86%
Together AI100.00%
wafer100.00%
zai100.00%

Throughput

last hour

P50 output tokens per second.

Together AI253 tps
inceptron235 tps
nebius204 tps
friendli172 tps
parasail167 tps
wafer162 tps
runware132 tps
novita101 tps
gmicloud100 tps
digitalocean97 tps
alibaba87 tps
Baseten83 tps
DeepInfra75 tps
zai72 tps
streamlake71 tps
tencent50 tps

Latency

last hour

P50 time to first token.

friendli0.2 s
Together AI0.2 s
wafer0.3 s
parasail0.3 s
nebius0.5 s
inceptron0.5 s
digitalocean0.5 s
zai1.1 s
streamlake1.2 s
DeepInfra1.2 s
alibaba1.4 s
tencent1.4 s
runware1.5 s
Baseten1.5 s
novita2.0 s
gmicloud2.3 s

Getting started

Create a key on the API Keys page and set it as IB_API_KEY. See the quickstart.

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.inferencebureau.com/v1",
  apiKey: process.env.IB_API_KEY,
});

const res = await client.chat.completions.create({
  model: "zai/glm-5.2",
  messages: [{ role: "user", content: "Why is the sky blue?" }],
});

console.log(res.choices[0].message.content);

Top-level parameters

Standard Chat Completions parameters work as documented. These have model-specific behavior.

ParameterTypeRequiredDescription
modelstringYesModel ID in the form creator/model, e.g. zai/glm-5.2. Must be allowed by your team's Directives.
max_tokensnumberNoHard cap on generated tokens, and the basis for the balance reserved per request. GLM 5.2 supports up to 128,000 output tokens.
reasoning{ effort: 'low' | 'medium' | 'high' }NoReasoning effort, mapped to the provider's native configuration. Reasoning tokens count toward output usage.
toolsTool[]NoFunction definitions the model can call. Tool calls are returned in choices[].message.tool_calls.
Supportedmax_tokenstemperaturestoptoolstool_choicereasoninginclude_reasoningresponse_formatstructured_outputs

Input limits

InputSourcesLimits
Text—Prompt and response share the 1M-token context window

More models by zai

ModelContextInputOutputCacheRelease
zai/glm-5.3-flashx-legal1M$0.37$1.25Read $0.075Write —11/02/2026
zai/glm-5.3-flashx1M$0.37$1.25Read $0.075Write —09/18/2026
zai/glm-5.3-fast1.05M$2.10$6.60Read $0.21Write —09/02/2026
zai/glm-5.3-flash1M$0.15$0.50Read $0.030Write —08/26/2026
zai/glm-5.31M$1.40$4.40Read $0.14Write —08/18/2026
zai/glm-5.2-fast1M$2.80$8.80Read $0.56Write —06/23/2026
zai/glm-5.1203K$1.40$4.40Read $0.26Write —04/07/2026
zai/glm-5v-turbo200K$1.20$4.00Read $0.24Write —04/01/2026
zai/glm-5-turbo203K$1.20$4.00Read $0.24Write —03/15/2026
zai/glm-5203K$1.00$3.20Read —Write —02/12/2026