gpt-5.6-luna

7/11/2026, 8:14:09 PM · 15 tests · run 61842902-5e2e-4cc1-b05b-7a06d2f2d11b

basic-generation

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

Produced 99 words (133 output tokens).

Result

{
  "wordCount": 99,
  "outputTokens": 133
}

adc88fd0-3f78-499f-823f-0e95eeefcd6e

Jul 11, 2026, 20:13:43

openai:gpt-5.6-luna

Input

Output

133

Reasoning

Answer

133

Cost

$0.000815

Mode

streaming

Cost breakdown

Input $0.000017Output $0.000798

Latency

1.8s

TTFT

650ms

Tok/s

73.0

Tok/s (adj)

113.6

usage-cost

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

in=14 out=9 rt=0 cost=$0.000068 (paid).

Result

{
  "inputTokens": 14,
  "outputTokens": 9,
  "reasoningTokens": 0,
  "costUsd": 0.000068,
  "paid": true
}

a843d5d8-8656-4e5e-86f9-0b6d7bd45f2f

Jul 11, 2026, 20:13:44

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000068

Mode

streaming

Cost breakdown

Input $0.000014Output $0.000054

Latency

647ms

TTFT

458ms

Tok/s

13.9

Tok/s (adj)

47.6

streaming-structure

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

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

Result

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

1b77ecea-341a-443c-a7f9-db7170b55531

Jul 11, 2026, 20:13:44

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000217

Mode

streaming

Cost breakdown

Input $0.000019Output $0.000198

Latency

728ms

TTFT

396ms

Tok/s

45.3

Tok/s (adj)

99.3

complete-stream-parity

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

stream="Paris" · complete="Paris"

Result

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

caad54bc-97cf-4f53-948a-114163726ad9

Jul 11, 2026, 20:13:45

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000050

Mode

streaming

Cost breakdown

Input $0.000020Output $0.000030

Latency

722ms

TTFT

555ms

Tok/s

6.9

Tok/s (adj)

29.9

b737f721-3481-4ce2-a322-d437c1cb8b4a

Jul 11, 2026, 20:13:46

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000050

Mode

non-streaming

Cost breakdown

Input $0.000020Output $0.000030

Latency

925ms

Tok/s

5.4

reasoning-conformance

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

Class (from reasoning-config): toggleable.
off: out=263 · mid: out=473 (rt=223).

Result

{
  "class": "toggleable",
  "mode": "deterministic",
  "offOutputTokens": 263,
  "midOutputTokens": 473,
  "midReasoningTokens": 223
}

7c2a36ee-3531-4be5-a5aa-f3d0a94855b0

Jul 11, 2026, 20:13:48

openai:gpt-5.6-luna

Input

Output

263

Reasoning

Answer

263

Cost

$0.0017

Mode

streaming

Cost breakdown

Input $0.000091Output $0.0016

Latency

1.6s

TTFT

413ms

Tok/s

161.8

Tok/s (adj)

217.1

ac001c98-581a-48b6-a79a-b9c1b16c16ed

Jul 11, 2026, 20:13:52

openai:gpt-5.6-luna

Input

Output

473

Reasoning

223

Answer

250

Cost

$0.0029

Mode

streaming

Cost breakdown

Input $0.000091Output $0.0028

Latency

4.2s

TTFT

3.6s

Tok/s

112.0

Tok/s (adj)

415.2

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
}

2f6ed18b-1ad3-47c6-b836-9e3b2124460a

Jul 11, 2026, 20:13:53

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000211

Mode

streaming

Cost breakdown

Input $0.000085Output $0.000126

Latency

938ms

Tok/s

22.4

tool-loop

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

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

Result

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

c3619adc-42ad-486b-bc5c-c881e198221c

Jul 11, 2026, 20:13:54

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000214

Mode

streaming

Cost breakdown

Input $0.000088Output $0.000126

Latency

803ms

Tok/s

26.2

85de942f-4cca-4a75-8a8d-48fd118a0c1a

Jul 11, 2026, 20:13:54

openai:gpt-5.6-luna

Input

124

Output

Reasoning

Answer

Cost

$0.000202

Mode

streaming

Cost breakdown

Input $0.000124Output $0.000078

Latency

789ms

TTFT

513ms

Tok/s

16.5

Tok/s (adj)

47.1

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.

reasoned=true, openai token —; follow-up → "Original balance: **$1,234** Calculation after 7 years at 5% annual c"

Result

{
  "signatureType": "openai",
  "tokenCaptured": false,
  "reasoned": true,
  "answered": true,
  "continuation": "Original balance: **$1,234**\n\nCalculation after 7 years at 5% annual compound growth:\n\n\\[\n1234(1.05)^7 = \\$1,734.15\n\\]\n\nRounded to the neare"
}

76e87e7e-f852-4296-9ec7-b76ccdb98dcc

Jul 11, 2026, 20:13:57

openai:gpt-5.6-luna

Input

124

Output

Reasoning

Answer

Cost

$0.000322

Mode

streaming

Cost breakdown

Input $0.000124Output $0.000198

Latency

2.6s

Tok/s

12.7

676717d1-6b60-48bf-8c43-2cae0a85796c

Jul 11, 2026, 20:13:58

openai:gpt-5.6-luna

Input

164

Output

102

Reasoning

Answer

102

Cost

$0.000776

Mode

streaming

Cost breakdown

Input $0.000164Output $0.000612

Latency

991ms

TTFT

403ms

Tok/s

102.9

Tok/s (adj)

173.4

prompt-cache

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

implicit cache; cacheRead per attempt [10173]; hit=yes.

Result

{
  "cacheStyle": "implicit",
  "cacheReadPerAttempt": [
    10173
  ],
  "cacheReadTokens": 10173,
  "cacheHitCost": 0.0012843000000000002
}

d04f1b4e-fa83-4617-9fbd-3de448989db4

Jul 11, 2026, 20:13:59

openai:gpt-5.6-luna

Input

10,176

Output

Reasoning

Answer

Cost

$0.0104

Mode

non-streaming

Cost breakdown

Input $0.0102Output $0.000216

Latency

1.0s

Tok/s

35.0

51ee1d43-62dc-47b8-abe8-4eb53dc5dde1

Jul 11, 2026, 20:14:00

openai:gpt-5.6-luna

Input

10,176

Output

Reasoning

Answer

Cache Read

10,173

Cost

$0.0013

Mode

non-streaming

Cost breakdown

Input $0.000003Output $0.000264Cache read $0.0010

Latency

1.1s

Tok/s

38.8

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, United States"
  },
  "validJson": true,
  "missingKeys": [],
  "bareJson": true
}

6d3602ac-1080-4d76-8484-17c0a51777b7

Jul 11, 2026, 20:14:01

openai:gpt-5.6-luna

Input

Output

Reasoning

Answer

Cost

$0.000229

Mode

streaming

Cost breakdown

Input $0.000067Output $0.000162

Latency

936ms

TTFT

556ms

Tok/s

28.8

Tok/s (adj)

71.0

summarize-memex

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

Read ~10208 input tokens, wrote 576 words (734 output tokens) in 5487ms.

Result

{
  "wordCount": 576,
  "inputTokens": 10208,
  "outputTokens": 734,
  "latencyMs": 5487
}

4fcb453d-6f16-4373-99b1-2b5c27bd97fe

Jul 11, 2026, 20:14:07

openai:gpt-5.6-luna

Input

10,208

Output

734

Reasoning

Answer

734

Cost

$0.0146

Mode

streaming

Cost breakdown

Input $0.0102Output $0.0044

Latency

5.5s

TTFT

555ms

Tok/s

133.8

Tok/s (adj)

148.8

truncation

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

finishReason=length, out=100 tokens.

Result

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

d7aed02f-8921-41c5-8192-db434a1bd3a7

Jul 11, 2026, 20:14:08

openai:gpt-5.6-luna

Input

Output

100

Reasoning

Answer

100

Cost

$0.000623

Mode

streaming

Cost breakdown

Input $0.000023Output $0.000600

Latency

1.7s

TTFT

530ms

Tok/s

58.2

Tok/s (adj)

84.2

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"
}