Armand McQueen

basic-generation

ok

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

Produced 109 words (138 output tokens).

Result

{
  "wordCount": 109,
  "outputTokens": 138
}

011fe92f-06fd-44a3-b979-bee221d82470

Sep 5, 2026, 02:35:25

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

29

Output

138

Reasoning

0

Answer

138

Cost

$0.000030

Mode

streaming

Cost breakdown

Input $0.000002Output $0.000028

Latency

635ms

TTFT

230ms

Tok/s

217.3

Tok/s (adj)

340.9

usage-cost

ok

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

in=25 out=6 rt=0 cost=$0.000003 (paid).

Result

{
  "inputTokens": 25,
  "outputTokens": 6,
  "reasoningTokens": 0,
  "costUsd": 0.0000032000000000000007,
  "paid": true
}

427e0abd-b06d-408e-98e5-30dfe4e76296

Sep 5, 2026, 02:35:25

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

25

Output

6

Reasoning

0

Answer

6

Cost

$0.000003

Mode

streaming

Cost breakdown

Input $0.000002Output $0.000001

Latency

289ms

TTFT

266ms

Tok/s

20.8

Tok/s (adj)

257.8

streaming-structure

ok

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

mode=streaming, 12 stream events, 1 text block(s), 35 chars.

Result

{
  "mode": "streaming",
  "streamEventCount": 12,
  "textBlocks": 1,
  "textLen": 35
}

21828bcb-f5c1-4f69-8277-0c2271d065ff

Sep 5, 2026, 02:35:26

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

30

Output

36

Reasoning

0

Answer

36

Cost

$0.000010

Mode

streaming

Cost breakdown

Input $0.000002Output $0.000007

Latency

156ms

TTFT

116ms

Tok/s

230.8

Tok/s (adj)

905.9

complete-stream-parity

ok

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

stream="Paris" · complete="Paris"

Result

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

ba77bfd7-ceae-42b1-b7c7-eec8ffda7384

Sep 5, 2026, 02:35:26

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

30

Output

2

Reasoning

0

Answer

2

Cost

$0.000003

Mode

streaming

Cost breakdown

Input $0.000002Output $0.000000

Latency

137ms

TTFT

126ms

Tok/s

14.6

Tok/s (adj)

184.7

7bb7c25a-8a5a-42ac-b1a4-f3eea6bf4274

Sep 5, 2026, 02:35:26

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

30

Output

2

Reasoning

0

Answer

2

Cost

$0.000003

Mode

non-streaming

Cost breakdown

Input $0.000002Output $0.000000

Latency

132ms

Tok/s

15.2

reasoning-conformance

ok

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

Class (from reasoning-config): toggleable.
off: out=693 · mid: out=1311 (rt=0, rchars=2432).

Result

{
  "class": "toggleable",
  "mode": "deterministic",
  "offOutputTokens": 693,
  "midOutputTokens": 1311,
  "midReasoningTokens": 0
}

e311c5d9-7235-4309-9fde-1a0ca2a2ad5d

Sep 5, 2026, 02:35:27

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

108

Output

693

Reasoning

0

Answer

693

Cost

$0.000147

Mode

streaming

Cost breakdown

Input $0.000009Output $0.000139

Latency

1.2s

TTFT

133ms

Tok/s

565.7

Tok/s (adj)

634.4

221e7ae1-cf0e-4e17-bdbd-5c4d2b38560c

Sep 5, 2026, 02:35:29

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

108

Output

1,311

Reasoning

0

Answer

1,311

Cost

$0.000271

Mode

streaming

Cost breakdown

Input $0.000009Output $0.000262

Latency

2.0s

TTFT

1.4s

Tok/s

648.0

Tok/s (adj)

1973.1

single-tool-call

ok

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
}

164fdf69-8ef3-463e-8579-151e45de084d

Sep 5, 2026, 02:35:29

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

366

Output

35

Reasoning

0

Answer

35

Cost

$0.000036

Mode

streaming

Cost breakdown

Input $0.000029Output $0.000007

Latency

203ms

Tok/s

172.4

tool-loop

ok

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 equals **391**."

Result

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

0697a3b5-d21d-46aa-8da0-7362bd88ce53

Sep 5, 2026, 02:35:31

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

360

Output

1,024

Reasoning

0

Answer

1,024

Cost

$0.000234

Mode

streaming

Cost breakdown

Input $0.000029Output $0.000205

Latency

1.3s

TTFT

173ms

Tok/s

803.1

Tok/s (adj)

929.4

62244751-234f-498a-8f46-4452511f24fc

Sep 5, 2026, 02:35:31

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

1,739

Output

14

Reasoning

0

Answer

14

Cost

$0.000142

Mode

streaming

Cost breakdown

Input $0.000139Output $0.000003

Latency

230ms

TTFT

206ms

Tok/s

60.9

Tok/s (adj)

579.0

web-search

ok

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

ok

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 → "**Account balance:** $1,234 **Future value after 7 years at 5% annu"

Result

{
  "signatureType": "none",
  "tokenCaptured": false,
  "reasoned": false,
  "answered": true,
  "continuation": "**Account balance:** $1,234  \n\n**Future value after 7 years at 5% annual compound interest**\n\n\\[\n\\text{Future value}=P\\,(1+r)^{n}=1234\\;(1+0"
}

faf1703f-f785-47ce-b1c5-3f4c38eb99aa

Sep 5, 2026, 02:35:32

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

376

Output

98

Reasoning

0

Answer

98

Cost

$0.000050

Mode

streaming

Cost breakdown

Input $0.000030Output $0.000020

Latency

1.0s

Tok/s

95.1

63cb6f2c-ad29-4768-9032-a76889771e3a

Sep 5, 2026, 02:35:33

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

500

Output

191

Reasoning

0

Answer

191

Cost

$0.000078

Mode

streaming

Cost breakdown

Input $0.000040Output $0.000038

Latency

640ms

TTFT

228ms

Tok/s

298.4

Tok/s (adj)

463.3

prompt-cache

ok

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

ok

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

2c4fa09a-13ed-4ca9-a2ed-7e87da91d461

Sep 5, 2026, 02:35:33

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

76

Output

34

Reasoning

0

Answer

34

Cost

$0.000013

Mode

streaming

Cost breakdown

Input $0.000006Output $0.000007

Latency

201ms

TTFT

121ms

Tok/s

169.2

Tok/s (adj)

425.0

summarize-memex

ok

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

Read ~10385 input tokens, wrote 364 words (446 output tokens) in 1891ms.

Result

{
  "wordCount": 364,
  "inputTokens": 10385,
  "outputTokens": 446,
  "latencyMs": 1891
}

434887c5-01fc-460f-93ed-c5e4ffc59310

Sep 5, 2026, 02:35:35

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

10,385

Output

446

Reasoning

0

Answer

446

Cost

$0.000920

Mode

streaming

Cost breakdown

Input $0.000831Output $0.000089

Latency

1.9s

TTFT

319ms

Tok/s

235.9

Tok/s (adj)

283.7

truncation

ok

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

finishReason=length, out=100 tokens.

Result

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

56ffe605-8f52-4cbf-a3f8-d6cc39186c90

Sep 5, 2026, 02:35:35

deepinfra:nvidia/NVIDIA-Nemotron-3.5-Lightning

Input

35

Output

100

Reasoning

0

Answer

100

Cost

$0.000023

Mode

streaming

Cost breakdown

Input $0.000003Output $0.000020

Latency

433ms

TTFT

111ms

Tok/s

230.9

Tok/s (adj)

310.8

streaming-only-rejection

ok

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

streamingOnly=false; complete() succeeded.

Result

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

abort

ok

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

pre-aborted signal → rejected (AbortError).

Result

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