gpt-5.6-terra

7/11/2026, 8:14:54 PM · 15 tests · run 2f37a35a-0b31-4f55-81ab-adc3135039de

basic-generation

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

Produced 94 words (132 output tokens).

Result

{
  "wordCount": 94,
  "outputTokens": 132
}

5182529c-d51d-4069-9f26-7dc5def8fbcf

Jul 11, 2026, 20:14:20

openai:gpt-5.6-terra

Input

Output

132

Reasoning

Answer

132

Cost

$0.0020

Mode

streaming

Cost breakdown

Input $0.000043Output $0.0020

Latency

1.8s

TTFT

588ms

Tok/s

72.4

Tok/s (adj)

106.8

usage-cost

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

in=14 out=9 rt=0 cost=$0.000170 (paid).

Result

{
  "inputTokens": 14,
  "outputTokens": 9,
  "reasoningTokens": 0,
  "costUsd": 0.00017,
  "paid": true
}

192ea555-1355-43f7-95d6-0ad748b776d3

Jul 11, 2026, 20:14:21

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000170

Mode

streaming

Cost breakdown

Input $0.000035Output $0.000135

Latency

945ms

TTFT

705ms

Tok/s

9.5

Tok/s (adj)

37.4

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
}

0acbae7d-a9ec-4abf-bc43-017ed03a1225

Jul 11, 2026, 20:14:22

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000543

Mode

streaming

Cost breakdown

Input $0.000048Output $0.000495

Latency

1.1s

TTFT

623ms

Tok/s

31.1

Tok/s (adj)

75.1

complete-stream-parity

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

stream="Paris" · complete="Paris"

Result

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

e945f781-1199-4e83-8a02-c5b29fc16f68

Jul 11, 2026, 20:14:23

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000125

Mode

streaming

Cost breakdown

Input $0.000050Output $0.000075

Latency

901ms

TTFT

720ms

Tok/s

5.5

Tok/s (adj)

27.7

98f5c3d2-ca66-46d8-80e8-160de41c50ee

Jul 11, 2026, 20:14:24

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000125

Mode

non-streaming

Cost breakdown

Input $0.000050Output $0.000075

Latency

923ms

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=255 · mid: out=307 (rt=74).

Result

{
  "class": "toggleable",
  "mode": "deterministic",
  "offOutputTokens": 255,
  "midOutputTokens": 307,
  "midReasoningTokens": 74
}

267575b2-4c6e-4889-afc4-655028965e8e

Jul 11, 2026, 20:14:26

openai:gpt-5.6-terra

Input

Output

255

Reasoning

Answer

255

Cost

$0.0041

Mode

streaming

Cost breakdown

Input $0.000228Output $0.0038

Latency

2.4s

TTFT

545ms

Tok/s

108.3

Tok/s (adj)

141.0

5522e031-b372-4a32-842c-7f8222de1dd4

Jul 11, 2026, 20:14:31

openai:gpt-5.6-terra

Input

Output

307

Reasoning

Answer

233

Cost

$0.0048

Mode

streaming

Cost breakdown

Input $0.000228Output $0.0046

Latency

5.3s

TTFT

4.9s

Tok/s

57.4

Tok/s (adj)

476.8

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
}

6dd9b61d-6fcf-4719-8c7a-c0275ceec8f3

Jul 11, 2026, 20:14:32

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000527

Mode

streaming

Cost breakdown

Input $0.000213Output $0.000315

Latency

805ms

Tok/s

26.1

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
}

27904df3-041d-463e-b9e1-9d26d3278ace

Jul 11, 2026, 20:14:33

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000535

Mode

streaming

Cost breakdown

Input $0.000220Output $0.000315

Latency

963ms

Tok/s

21.8

a7f35e87-4132-464b-97a6-4e90118bf48a

Jul 11, 2026, 20:14:34

openai:gpt-5.6-terra

Input

124

Output

Reasoning

Answer

Cost

$0.000505

Mode

streaming

Cost breakdown

Input $0.000310Output $0.000195

Latency

777ms

TTFT

450ms

Tok/s

16.7

Tok/s (adj)

39.7

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 → "Current balance: **$1,234** Calculation after 7 years at 5% annual co"

Result

{
  "signatureType": "openai",
  "tokenCaptured": false,
  "reasoned": true,
  "answered": true,
  "continuation": "Current balance: **$1,234**\n\nCalculation after 7 years at 5% annual compound growth:\n\n\\[\n1234 \\times (1.05)^7 = 1234 \\times 1.40710042265625"
}

ffbc0acf-0af4-43f6-bbc9-fcc3a38d104f

Jul 11, 2026, 20:14:35

openai:gpt-5.6-terra

Input

124

Output

Reasoning

Answer

Cost

$0.000790

Mode

streaming

Cost breakdown

Input $0.000310Output $0.000480

Latency

1.2s

Tok/s

27.7

2442c390-262e-4ce3-ba41-063a14644e21

Jul 11, 2026, 20:14:37

openai:gpt-5.6-terra

Input

164

Output

114

Reasoning

Answer

114

Cost

$0.0021

Mode

streaming

Cost breakdown

Input $0.000410Output $0.0017

Latency

1.5s

TTFT

632ms

Tok/s

75.0

Tok/s (adj)

128.2

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": 0.00316575
}

3faf0116-0379-49a3-941e-9f8619fad1ca

Jul 11, 2026, 20:14:38

openai:gpt-5.6-terra

Input

10,176

Output

Reasoning

Answer

Cost

$0.0260

Mode

non-streaming

Cost breakdown

Input $0.0254Output $0.000600

Latency

1.6s

Tok/s

24.7

979ab07a-d4ad-4b08-8f2e-df659565c3ab

Jul 11, 2026, 20:14:42

openai:gpt-5.6-terra

Input

10,176

Output

Reasoning

Answer

Cache Read

10,173

Cost

$0.0032

Mode

non-streaming

Cost breakdown

Input $0.000008Output $0.000615Cache read $0.0025

Latency

3.5s

Tok/s

11.9

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
}

885221f3-7413-4695-b9fe-1d52dd2f1e26

Jul 11, 2026, 20:14:42

openai:gpt-5.6-terra

Input

Output

Reasoning

Answer

Cost

$0.000572

Mode

streaming

Cost breakdown

Input $0.000168Output $0.000405

Latency

729ms

TTFT

440ms

Tok/s

37.0

Tok/s (adj)

93.3

summarize-memex

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

Read ~10208 input tokens, wrote 625 words (794 output tokens) in 8755ms.

Result

{
  "wordCount": 625,
  "inputTokens": 10208,
  "outputTokens": 794,
  "latencyMs": 8755
}

d9cc6cbd-6384-44b7-a7fb-3f3c6068ae1f

Jul 11, 2026, 20:14:51

openai:gpt-5.6-terra

Input

10,208

Output

794

Reasoning

Answer

794

Cost

$0.0374

Mode

streaming

Cost breakdown

Input $0.0255Output $0.0119

Latency

8.8s

TTFT

797ms

Tok/s

90.7

Tok/s (adj)

99.8

truncation

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

finishReason=length, out=100 tokens.

Result

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

f4e3b127-593b-496a-b9bf-1cf75cf2f618

Jul 11, 2026, 20:14:53

openai:gpt-5.6-terra

Input

Output

100

Reasoning

Answer

100

Cost

$0.0016

Mode

streaming

Cost breakdown

Input $0.000058Output $0.0015

Latency

2.0s

TTFT

605ms

Tok/s

50.9

Tok/s (adj)

73.6

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"
}