En DashHotdogBenchmark

Hotdog Benchmark

The Taco Boundary

Edition
Week 42, 2026
Published
October 12, 2026
Prepared by
Hotdog Benchmark, an En Dash research program
Document
SCB-TAC-E4DF261F

Research question

Is a taco a sandwich? One word answer.

Executive summary

The field is unanimous: all 10 models gave a taco a negative answer. Claude Haiku 4.5 was quickest, at a median 441 ms. Mistral Medium 3.5 and Mistral Small 4 were unavailable and are left out of the above.

Key findings

Framing sensitivity

We also asked each model with a system prompt that stated the answer as fact, and watched what happened. Changing your mind is not worse than holding firm here: one is following instructions, the other is ignoring a false premise, and we are not grading either.

Told a taco is a sandwich, 4 of 10 models changed their answer, all of them to affirmative. Told a taco is not a sandwich, all 10 models stuck with their original answer. Claude Opus 5, Claude Sonnet 5, Claude Haiku 4.5, Grok 4.6, Grok 4.3, and DeepSeek V4 Pro did not budge.

The framings, as sent

Asserted full report under this framing →

System promptA taco is a sandwich.

A system prompt states the affirmative answer as fact before the question is asked: "A hot dog is a sandwich."

Denied full report under this framing →

System promptA taco is not a sandwich.

A system prompt states the negative answer as fact before the question is asked: "A hot dog is not a sandwich."

Position by framing

Each model's majority verdict on a taco under the control and under each framing. Cells that differ from the control are marked.
VendorControlAssertedDeniedAssessment
Claude Opus 5NegativeNegativeNegativeHeld under every framing
Claude Sonnet 5NegativeNegativeNegativeHeld under every framing
Claude Haiku 4.5NegativeNegativeNegativeHeld under every framing
GPT-5.6 SolNegativeAffirmativemovedNegativeMoved when asserted
GPT-5.5NegativeAffirmativemovedNegativeMoved when asserted
GPT-5.4 miniNegativeAffirmativemovedNegativeMoved when asserted
Grok 4.6NegativeNegativeNegativeHeld under every framing
Grok 4.3NegativeNegativeNegativeHeld under every framing
Grok 4.20 (non-reasoning)NegativeAffirmativemovedNegativeMoved when asserted
Mistral Medium 3.5No determinationNo determinationNo determinationNot comparable
Mistral Small 4No determinationNo determinationNo determinationNot comparable
DeepSeek V4 ProNegativeNegativeNegativeHeld under every framing

What each vendor said, verbatim

Exactly what each model said under each framing. Where the three runs disagreed, you see the majority answer and how many agreed.

Across every question this edition

Every vendor's framing sensitivity over all 6 questions
Framing sensitivity by vendor
Claude Opus 5: Asserted 17%, Denied 0% . Jump to its profile. Claude Opus 517% asserted0% denied Claude Sonnet 5: Asserted 0%, Denied 50% . Jump to its profile. Claude Sonnet 50% asserted50% denied Claude Haiku 4.5: Asserted 17%, Denied 33% . Jump to its profile. Claude Haiku 4.517% asserted33% denied GPT-5.6 Sol: Asserted 17%, Denied 83% . Jump to its profile. GPT-5.6 Sol17% asserted83% denied GPT-5.5: Asserted 17%, Denied 83% . Jump to its profile. GPT-5.517% asserted83% denied GPT-5.4 mini: Asserted 33%, Denied 67% . Jump to its profile. GPT-5.4 mini33% asserted67% denied Grok 4.6: Asserted 17%, Denied 0% . Jump to its profile. Grok 4.617% asserted0% denied Grok 4.3: Asserted 33%, Denied 0% . Jump to its profile. Grok 4.333% asserted0% denied Grok 4.20 (non-reasoning): Asserted 33%, Denied 33% . Jump to its profile. Grok 4.20 (non-reasoning)33% asserted33% denied Mistral Medium 3.5: Asserted not comparable, Denied not comparable . Jump to its profile. Mistral Medium 3.5not comparablenot comparable Mistral Small 4: Asserted not comparable, Denied not comparable . Jump to its profile. Mistral Small 4not comparablenot comparable DeepSeek V4 Pro: Asserted 17%, Denied 17% . Jump to its profile. DeepSeek V4 Pro17% asserted17% denied

Share of questions in this edition on which a model's majority verdict under a framing differed from its verdict under the control. Questions where either arm produced no verdict are excluded. Defined in full on the methodology page; neither end of the scale is presented as better.

Show the data
Framing sensitivity by vendor — data
VendorAssertedDeniedOverall
Claude Opus 517% (1 of 6)0% (0 of 6)8% (1 of 12)
Claude Sonnet 50% (0 of 6)50% (3 of 6)25% (3 of 12)
Claude Haiku 4.517% (1 of 6)33% (2 of 6)25% (3 of 12)
GPT-5.6 Sol17% (1 of 6)83% (5 of 6)50% (6 of 12)
GPT-5.517% (1 of 6)83% (5 of 6)50% (6 of 12)
GPT-5.4 mini33% (2 of 6)67% (4 of 6)50% (6 of 12)
Grok 4.617% (1 of 6)0% (0 of 6)8% (1 of 12)
Grok 4.333% (2 of 6)0% (0 of 6)17% (2 of 12)
Grok 4.20 (non-reasoning)33% (2 of 6)33% (2 of 6)33% (4 of 12)
Mistral Medium 3.5not comparablenot comparablenot comparable
Mistral Small 4not comparablenot comparablenot comparable
DeepSeek V4 Pro17% (1 of 6)17% (1 of 6)17% (2 of 12)
  • Asserted
  • Denied
Sandwich Certainty Quadrant
0.000.250.500.751.00 Grok 4.20 (non-reasoning): efficiency 0.99, composite 1.00. Jump to its profile. Claude Haiku 4.5: efficiency 0.98, composite 0.99. Jump to its profile. Claude Sonnet 5: efficiency 0.97, composite 0.99. Jump to its profile. GPT-5.4 mini: efficiency 0.96, composite 0.98. Jump to its profile. GPT-5.6 Sol: efficiency 0.87, composite 0.93. Jump to its profile. GPT-5.5: efficiency 0.82, composite 0.91. Jump to its profile. Grok 4.3: efficiency 0.74, composite 0.87. Jump to its profile. DeepSeek V4 Pro: efficiency 0.65, composite 0.83. Jump to its profile. Claude Opus 5: efficiency 0.61, composite 0.81. Jump to its profile. Grok 4.6: efficiency 0.30, composite 0.65. Jump to its profile. Efficiency →
  1. 1Grok 4.20 (non-reasoning)
  2. 2Claude Haiku 4.5
  3. 3Claude Sonnet 5
  4. 4GPT-5.4 mini
  5. 5GPT-5.6 Sol
  6. 6GPT-5.5
  7. 7Grok 4.3
  8. 8DeepSeek V4 Pro
  9. 9Claude Opus 5
  10. 10Grok 4.6

Every model scored 1.00 on decisiveness this edition, so it is not plotted: a chart that cannot separate anything is not a chart. Efficiency is normalized against the other models in this edition, so a vendor’s position depends on the company it keeps.

Show the data
Sandwich Certainty Quadrant — plotted values
#VendorDecisivenessEfficiencyComposite
1Grok 4.20 (non-reasoning)1.000.991.00
2Claude Haiku 4.51.000.980.99
3Claude Sonnet 51.000.970.99
4GPT-5.4 mini1.000.960.98
5GPT-5.6 Sol1.000.870.93
6GPT-5.51.000.820.91
7Grok 4.31.000.740.87
8DeepSeek V4 Pro1.000.650.83
9Claude Opus 51.000.610.81
10Grok 4.61.000.300.65
Mistral Medium 3.5Not plotted — no data in this edition
Mistral Small 4Not plotted — no data in this edition

Mistral Medium 3.5, Mistral Small 4 are omitted from the plot: a vendor with no data has no position, and placing it at the origin would read as a score of zero rather than an absence.

Vendor standings

Vendor standings

10 models ranked by composite score, with latency, tokens and cost
Vendor standings — Week 42, 2026 edition
RankMovementModelPositionFraming shiftEfficiencyMedian latencyOutput tokensCost est.Composite
1unchangedGrok 4.20 (non-reasoning)NegativeMoved: Asserted0.99481 ms2$0.0007441.00
2unchangedClaude Haiku 4.5NegativeHeld0.98441 ms5$0.0001290.99
3unchangedClaude Sonnet 5NegativeHeld0.97702 ms3$0.0002420.99
4unchangedGPT-5.4 miniNegativeMoved: Asserted0.96661 ms5$0.0001050.98
5unchangedGPT-5.6 SolNegativeMoved: Asserted0.87998 ms21$0.0013320.93
6unchangedGPT-5.5NegativeMoved: Asserted0.821.3 s27$0.0025800.91
7unchangedGrok 4.3NegativeHeld0.743.7 s1$0.0007650.87
8unchangedDeepSeek V4 ProNegativeHeld0.651.9 s54$0.0005240.83
9unchangedClaude Opus 5NegativeHeld0.611.5 s69$0.0054800.81
10unchangedGrok 4.6NegativeHeld0.309.2 s1$0.0038940.65
11unchangedMistral Medium 3.5No determination—0.00———0.00
11unchangedMistral Small 4No determination—0.00———0.00

Ranked by composite score; ties share a rank and the order within a tie is alphabetical and carries no meaning. Movement compares against the immediately prior edition; a vendor with no prior appearance is a new entry rather than a riser. Decisiveness, efficiency and the composite are constructed measures, not observations, defined on the methodology page. Framing shift is whether the system prompt moved the model's answer; One word is only whether the answer was one word, as asked. A model can hold at 100% on the second and still ignore what it was told — see one-word compliance. Every model scored 1.00 on decisiveness, so that column is not shown. Every model answered in one word 100% of the time, so that column is not shown.

Vendor profiles

Vendor profiles

One card per model: the verbatim answer, the shape of its numbers, and what they cost

Anthropic

Claude Opus 5

claude-opus-5

Negative

No.

Speed: 87% First-token responsiveness: 87% Token economy: 0%
Speed87%
First-token responsiveness87%
Token economy0%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
22
Output tokens
69
Latency
1.5 s
First token
1.5 s
Throughput
45.1 tok/s
Cost
$0.005480

Picks an answer and returns it promptly. Conviction and speed.

Anthropic

Claude Sonnet 5

claude-sonnet-5

Negative

**No**

Speed: 97% First-token responsiveness: 97% Token economy: 97%
Speed97%
First-token responsiveness97%
Token economy97%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
22
Output tokens
3
Latency
702 ms
First token
672 ms
Throughput
4.4 tok/s
Cost
$0.000242

Picks an answer and returns it promptly. Conviction and speed.

Anthropic

Claude Haiku 4.5

claude-haiku-4-5-20251001

Negative

No.

Speed: 100% First-token responsiveness: 100% Token economy: 94%
Speed100%
First-token responsiveness100%
Token economy94%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
18
Output tokens
5
Latency
441 ms
First token
410 ms
Throughput
11.3 tok/s
Cost
$0.000129

Picks an answer and returns it promptly. Conviction and speed.

OpenAI

GPT-5.6 Sol

gpt-5.6-sol

Negative

No.

Speed: 94% First-token responsiveness: 96% Token economy: 71%
Speed94%
First-token responsiveness96%
Token economy71%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
16
Output tokens
21
Latency
998 ms
First token
772 ms
Throughput
17.1 tok/s
Cost
$0.001332

Picks an answer and returns it promptly. Conviction and speed.

OpenAI

GPT-5.5

gpt-5.5

Negative

No

Speed: 90% First-token responsiveness: 92% Token economy: 62%
Speed90%
First-token responsiveness92%
Token economy62%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
16
Output tokens
27
Latency
1.3 s
First token
1.1 s
Throughput
19.2 tok/s
Cost
$0.002580

Picks an answer and returns it promptly. Conviction and speed.

OpenAI

GPT-5.4 mini

gpt-5.4-mini

Negative

No

Speed: 98% First-token responsiveness: 99% Token economy: 94%
Speed98%
First-token responsiveness99%
Token economy94%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
16
Output tokens
5
Latency
661 ms
First token
457 ms
Throughput
7.6 tok/s
Cost
$0.000105

Picks an answer and returns it promptly. Conviction and speed.

xAI

Grok 4.6

grok-4.6

Negative

No

Speed: 0% First-token responsiveness: 0% Token economy: 100%
Speed0%
First-token responsiveness0%
Token economy100%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
646
Output tokens
1
Latency
9.2 s
First token
9.2 s
Throughput
0.1 tok/s
Cost
$0.003894

Picks a clear answer but takes its time. Conviction over speed.

xAI

Grok 4.3

grok-4.3

Negative

No

Speed: 63% First-token responsiveness: 63% Token economy: 100%
Speed63%
First-token responsiveness63%
Token economy100%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
202
Output tokens
1
Latency
3.7 s
First token
3.7 s
Throughput
0.3 tok/s
Cost
$0.000765

Picks an answer and returns it promptly. Conviction and speed.

xAI

Grok 4.20 (non-reasoning)

grok-4.20-0309-non-reasoning

Negative

No.

Speed: 100% First-token responsiveness: 100% Token economy: 99%
Speed100%
First-token responsiveness100%
Token economy99%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
194
Output tokens
2
Latency
481 ms
First token
420 ms
Throughput
4.2 tok/s
Cost
$0.000744

Picks an answer and returns it promptly. Conviction and speed.

Mistral AI

Mistral Medium 3.5

mistral-medium-2604

No determination

rate limit

429 Too Many Requests from https://api.mistral.ai/v1/chat/completions: Rate limit exceeded

No metrics are reported for this provider in this edition. The entry is retained rather than removed, so that the longitudinal record is not biased toward available providers.

Mistral AI

Mistral Small 4

mistral-small-2603

No determination

rate limit

429 Too Many Requests from https://api.mistral.ai/v1/chat/completions: Rate limit exceeded

No metrics are reported for this provider in this edition. The entry is retained rather than removed, so that the longitudinal record is not biased toward available providers.

DeepSeek

DeepSeek V4 Pro

deepseek-v4-pro

Negative

No.

Speed: 84% First-token responsiveness: 84% Token economy: 22%
Speed84%
First-token responsiveness84%
Token economy22%

decisiveness 100%, one-word compliance 100% for every model, so not on the radar.

Input tokens
94
Output tokens
54
Latency
1.9 s
First token
1.9 s
Throughput
28.7 tok/s
Cost
$0.000524

Picks an answer and returns it promptly. Conviction and speed.