deepinfra:glm-5.3

9/5/2026, 2:22:24 AM · 15 tests · run add0ceab-0caf-4c86-973c-668064e7210a

basic-generation

Plain streaming generation with a soft length constraint (~100 words).

Produced 98 words (300 output tokens).

Result

{
  "wordCount": 98,
  "outputTokens": 300
}

5bd08b2b-cb85-4fbd-ac31-c26d16edc4e2

Sep 5, 2026, 02:21:15

deepinfra:zai-org/GLM-5.3

Input

Output

300

Reasoning

Answer

300

Cost

$0.0012

Mode

streaming

Cost breakdown

Input $0.000028Output $0.0012

Latency

11.0s

TTFT

8.8s

Tok/s

27.2

Tok/s (adj)

136.7

usage-cost

Asserts usage tokens populate and cost computes from the ChatReq contract.

in=20 out=272 rt=0 cost=$0.001112 (paid).

Result

{
  "inputTokens": 20,
  "outputTokens": 272,
  "reasoningTokens": 0,
  "costUsd": 0.001112,
  "paid": true
}

bc73f24b-0139-41fb-934d-f144546071ca

Sep 5, 2026, 02:21:19

deepinfra:zai-org/GLM-5.3

Input

Output

272

Reasoning

Answer

272

Cost

$0.0011

Mode

streaming

Cost breakdown

Input $0.000024Output $0.0011

Latency

3.6s

TTFT

3.6s

Tok/s

75.3

Tok/s (adj)

335802.5

streaming-structure

Asserts a streaming call captures events that reconstruct into the final text.

mode=streaming, 21 stream events, 1 text block(s), 35 chars.

Result

{
  "mode": "streaming",
  "streamEventCount": 21,
  "textBlocks": 1,
  "textLen": 35
}

c9f1d089-6db6-425c-8ce2-aa0d463d9a94

Sep 5, 2026, 02:21:20

deepinfra:zai-org/GLM-5.3

Input

Output

Reasoning

Answer

Cost

$0.000286

Mode

streaming

Cost breakdown

Input $0.000030Output $0.000256

Latency

855ms

TTFT

704ms

Tok/s

74.9

Tok/s (adj)

424.5

complete-stream-parity

Same prompt via stream() and complete() both return valid, matching results.

stream="Paris" · complete="Paris"

Result

{
  "streamText": "Paris",
  "completeText": "Paris"
}

cc2510b8-4c7f-48cc-8279-507a60d2243b

Sep 5, 2026, 02:21:20

deepinfra:zai-org/GLM-5.3

Input

Output

Reasoning

Answer

Cost

$0.000043

Mode

streaming

Cost breakdown

Input $0.000031Output $0.000012

Latency

493ms

TTFT

491ms

Tok/s

6.1

Tok/s (adj)

1327.4

9570c75b-8dc0-47ba-bdac-c8ae6b800372

Sep 5, 2026, 02:21:23

deepinfra:zai-org/GLM-5.3

Input

Output

Reasoning

Answer

Cost

$0.000043

Mode

non-streaming

Cost breakdown

Input $0.000031Output $0.000012

Latency

2.2s

Tok/s

1.4

reasoning-conformance

Asserts the model's reasoning knob behaves as its reasoning-config declares.

Class (from reasoning-config): always-on.
reasoningTokens=0, outputTokens=447.

Result

{
  "class": "always-on",
  "reasoningTokens": 0,
  "outputTokens": 447
}

7f139d13-bbb8-49d7-affb-8d15008f6c79

Sep 5, 2026, 02:21:34

deepinfra:zai-org/GLM-5.3

Input

Output

447

Reasoning

Answer

447

Cost

$0.0019

Mode

streaming

Cost breakdown

Input $0.000116Output $0.0018

Latency

11.1s

TTFT

7.6s

Tok/s

40.3

Tok/s (adj)

126.5

single-tool-call

Offers a calculator tool and checks the model calls it for an arithmetic question.

Called "multiply" with args {"a":23,"b":17}.

Result

{
  "toolCalled": true,
  "toolName": "multiply",
  "args": {
    "a": 23,
    "b": 17
  },
  "correct": true
}

145ae021-26e3-4fc3-ac66-2e5c9449c290

Sep 5, 2026, 02:21:34

deepinfra:zai-org/GLM-5.3

Input

230

Output

Reasoning

Answer

Cost

$0.000344

Mode

streaming

Cost breakdown

Input $0.000276Output $0.000068

Latency

508ms

Tok/s

33.5

tool-loop

Forces a tool call, feeds the result back, and checks the continuation uses it.

step1 (required): multiply({"a":23,"b":17}) → step2: "The product of 23 multiplied by 17 is **391**."

Result

{
  "toolName": "multiply",
  "args": {
    "a": 23,
    "b": 17
  },
  "product": 391,
  "continuation": "The product of 23 multiplied by 17 is **391**.",
  "usesResult": true,
  "forced": true
}

79909491-c92d-4726-97e5-df2ce1d6e48f

Sep 5, 2026, 02:21:45

deepinfra:zai-org/GLM-5.3

Input

226

Output

1,024

Reasoning

Answer

1,024

Cost

$0.0044

Mode

streaming

Cost breakdown

Input $0.000271Output $0.0041

Latency

10.9s

Tok/s

93.8

2cb36e35-f733-4dd0-8b06-d42ffc6672ea

Sep 5, 2026, 02:21:49

deepinfra:zai-org/GLM-5.3

Input

1,761

Output

Reasoning

Answer

Cache Read

576

Cost

$0.0025

Mode

streaming

Cost breakdown

Input $0.0014Output $0.000380Cache read $0.000691

Latency

3.4s

TTFT

3.3s

Tok/s

27.6

Tok/s (adj)

874.8

web-search

Enables server-side web search and checks invocation counts, the per-request cost surcharge, and provider-executed output blocks.

Skipped: model does not support the catalog's web_search tool version.

Result

{
  "skipped": "no server-side web search support"
}

reasoning-tool-continuity

Reasoning model: continuity token captured on a tool call and replayed so the follow-up succeeds.

Step 1 didn't engage reasoning (adaptive discretion) — signature not exercised this run.
reasoned=false, none token —; follow-up → "**Current balance:** $1,234 **Projection (5% annual, compounded, 7 ye"

Result

{
  "signatureType": "none",
  "tokenCaptured": false,
  "reasoned": false,
  "answered": true,
  "continuation": "**Current balance:** $1,234\n\n**Projection (5% annual, compounded, 7 years):**\n\nBalance after 7 years = 1234 × 1.05⁷\n\n- 1.05⁷ = 1.40710042226"
}

2a9e052d-d5dc-4470-b7c0-ef35cb094ace

Sep 5, 2026, 02:21:53

deepinfra:zai-org/GLM-5.3

Input

253

Output

Reasoning

Answer

Cache Read

128

Cost

$0.000384

Mode

streaming

Cost breakdown

Input $0.000150Output $0.000080Cache read $0.000154

Latency

1.6s

Tok/s

12.5

4e5aa37c-98c7-4eae-b7c1-005b933a1536

Sep 5, 2026, 02:21:56

deepinfra:zai-org/GLM-5.3

Input

285

Output

304

Reasoning

Answer

304

Cost

$0.0016

Mode

streaming

Cost breakdown

Input $0.000342Output $0.0012

Latency

3.2s

TTFT

1.5s

Tok/s

94.4

Tok/s (adj)

178.3

prompt-cache

Repeat a large prefix; assert cache-read tokens are reported and cost drops.

Skipped: provider has no (known) prompt-cache reporting.

Result

{
  "skipped": "cacheStyle=none"
}

json-output

Requests a strict JSON object and checks it parses with the required keys.

Output parsed as JSON.
All required keys present.
Output is bare JSON (no leading prose/fence).

Result

{
  "parsed": {
    "name": "Anthropic",
    "founded": 2021,
    "headquarters": "San Francisco, California, USA"
  },
  "validJson": true,
  "missingKeys": [],
  "bareJson": true
}

5f2ee9e9-e78d-40b9-a916-3ebe92563e26

Sep 5, 2026, 02:21:56

deepinfra:zai-org/GLM-5.3

Input

Output

Reasoning

Answer

Cost

$0.000200

Mode

streaming

Cost breakdown

Input $0.000084Output $0.000116

Latency

428ms

TTFT

285ms

Tok/s

67.8

Tok/s (adj)

203.3

summarize-memex

Summarizes an ~11k-token essay to ~500 words (long input, reasoning minimized).

Read ~10238 input tokens, wrote 629 words (833 output tokens) in 24868ms.

Result

{
  "wordCount": 629,
  "inputTokens": 10238,
  "outputTokens": 833,
  "latencyMs": 24868
}

97ca4e88-0570-4664-9c67-c8c1527bd966

Sep 5, 2026, 02:22:21

deepinfra:zai-org/GLM-5.3

Input

10,238

Output

833

Reasoning

Answer

833

Cost

$0.0156

Mode

streaming

Cost breakdown

Input $0.0123Output $0.0033

Latency

24.9s

TTFT

2.8s

Tok/s

33.5

Tok/s (adj)

37.7

truncation

Tiny maxTokens on a long prompt yields finishReason:"length".

finishReason=length, out=100 tokens.

Result

{
  "finishReason": "length",
  "outputTokens": 100
}

c103de17-ec65-4c85-9bce-7bc1c90f4f41

Sep 5, 2026, 02:22:23

deepinfra:zai-org/GLM-5.3

Input

Output

100

Reasoning

Answer

100

Cost

$0.000435

Mode

streaming

Cost breakdown

Input $0.000035Output $0.000400

Latency

2.0s

TTFT

507ms

Tok/s

49.2

Tok/s (adj)

65.6

streaming-only-rejection

Streaming-only models reject complete(); all others accept it.

streamingOnly=false; complete() succeeded.

Result

{
  "streamingOnly": false,
  "completeThrew": false
}

abort

A pre-aborted signal rejects the call (abort is threaded + honored).

pre-aborted signal → rejected (AbortError).

Result

{
  "rejected": true,
  "errorName": "AbortError"
}