Claude Opus 5
| Decisiveness | 100% |
|---|---|
| Speed | 66% |
| First-token responsiveness | 65% |
| Token economy | 0% |
| Instruction compliance | 100% |
Picks a clear answer but takes its time. Conviction over speed.
Hotdog Benchmark
The Hamburger Control
Research question
Is a hamburger a sandwich? One word answer.
Key performance indicators
Models evaluated
11
Consensus position
Affirmative
100% of the field
Median latency
1.1 s
Median output tokens
5
Instruction compliance
100%
Executive summary
The field is unanimous: all 11 models gave a hamburger a affirmative answer. Mistral Small 4 was quickest, at a median 340 ms.
Key findings
- Consensus. Everyone agrees: a hamburger gets a affirmative from every model, unanimous. That does not happen often.
- Response latency. Mistral Small 4 answered in a median 340 ms; Grok 4.6 took 7.0 seconds. That is a 20.8× spread, mostly thinking time.
- Response length. Claude Opus 5 used the most output tokens, a median of 123, for a question that asked for one word.
- Instruction compliance. Every model kept it to one word. Nice.
- Composite standing. Mistral Small 4 tops the composite score at 1.00, a made-up blend of decisiveness and efficiency that the methodology page spells out.
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.
Changed position
7 of 11
comparable models, under at least one framing
Moved when asserted
0 of 11
0%
Moved when denied
7 of 11
64%
Told a hamburger is a sandwich, all 11 models stuck with their original answer. Told a hamburger is not a sandwich, 7 of 11 models changed their answer: Claude Sonnet 5 (affirmative to negative), GPT-5.6 Sol (affirmative to negative), GPT-5.5 (affirmative to negative), GPT-5.4 mini (affirmative to negative), Grok 4.20 (non-reasoning) (affirmative to negative), Mistral Medium 3.5 (affirmative to negative), and Mistral Small 4 (affirmative to negative). Claude Opus 5, Claude Haiku 4.5, Grok 4.6, and Grok 4.3 did not budge.
Show
Focusnone: pick a vendor in the table
The framings, as sent
- Asserted full report under this framing →
System prompt
A hamburger 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 prompt
A hamburger 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
| Vendor | Control | Asserted | Denied | Assessment |
|---|---|---|---|---|
| Claude Opus 5 | Affirmative | Affirmative | Affirmative | Held under every framing |
| Claude Sonnet 5 | Affirmative | Affirmative | Negativemoved | Moved when denied |
| Claude Haiku 4.5 | Affirmative | Affirmative | Affirmative | Held under every framing |
| GPT-5.6 Sol | Affirmative | Affirmative | Negativemoved | Moved when denied |
| GPT-5.5 | Affirmative | Affirmative | Negativemoved | Moved when denied |
| GPT-5.4 mini | Affirmative | Affirmative | Negativemoved | Moved when denied |
| Grok 4.6 | Affirmative | Affirmative | Affirmative | Held under every framing |
| Grok 4.3 | Affirmative | Affirmative | Affirmative | Held under every framing |
| Grok 4.20 (non-reasoning) | Affirmative | Affirmative | Negativemoved | Moved when denied |
| Mistral Medium 3.5 | Affirmative | Affirmative | Negativemoved | Moved when denied |
| Mistral Small 4 | Affirmative | Affirmative | Negativemoved | Moved when denied |
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.
Claude Opus 5held position
- Control no system prompt
Yes.
- Asserted told
A hamburger is a sandwich.
Yes.
- Denied told
A hamburger is not a sandwich.
Yes.
Claude Sonnet 5changed position
- Control no system prompt
**Yes**
- Asserted told
A hamburger is a sandwich.
**Yes**
- Denied told
A hamburger is not a sandwich.
No.
Claude Haiku 4.5held position
- Control no system prompt
Yes.
- Asserted told
A hamburger is a sandwich.
Yes.
- Denied told
A hamburger is not a sandwich.
Yes.
GPT-5.6 Solchanged position
- Control no system prompt
Yes.
- Asserted told
A hamburger is a sandwich.
Yes.
- Denied told
A hamburger is not a sandwich.
No.
GPT-5.5changed position
- Control no system prompt
Yes
- Asserted told
A hamburger is a sandwich.
Yes
- Denied told
A hamburger is not a sandwich.
No
GPT-5.4 minichanged position
- Control no system prompt
Yes
- Asserted told
A hamburger is a sandwich.
Yes
- Denied told
A hamburger is not a sandwich.
No
Grok 4.6held position
- Control no system prompt
Yes
- Asserted told
A hamburger is a sandwich.
Yes
- Denied told
A hamburger is not a sandwich.
Yes
Grok 4.3held position
- Control no system prompt
Yes
- Asserted told
A hamburger is a sandwich.
Yes
- Denied told
A hamburger is not a sandwich.
Yes
Grok 4.20 (non-reasoning)changed position
- Control no system prompt
Yes.
- Asserted told
A hamburger is a sandwich.
Yes
- Denied told
A hamburger is not a sandwich.
No
Mistral Medium 3.5changed position
- Control no system prompt
Yes.
- Asserted told
A hamburger is a sandwich.
Yes.
- Denied told
A hamburger is not a sandwich.
No
Mistral Small 4changed position
- Control no system prompt
Yes.
- Asserted told
A hamburger is a sandwich.
Yes
- Denied told
A hamburger is not a sandwich.
No
Across every question this edition
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
| Vendor | Asserted | Denied | Overall |
|---|---|---|---|
| Claude Opus 5 | 0% (0 of 3) | 0% (0 of 3) | 0% (0 of 6) |
| Claude Sonnet 5 | 0% (0 of 3) | 67% (2 of 3) | 33% (2 of 6) |
| Claude Haiku 4.5 | 0% (0 of 3) | 33% (1 of 3) | 17% (1 of 6) |
| GPT-5.6 Sol | 33% (1 of 3) | 67% (2 of 3) | 50% (3 of 6) |
| GPT-5.5 | 33% (1 of 3) | 67% (2 of 3) | 50% (3 of 6) |
| GPT-5.4 mini | 33% (1 of 3) | 67% (2 of 3) | 50% (3 of 6) |
| Grok 4.6 | 0% (0 of 3) | 0% (0 of 3) | 0% (0 of 6) |
| Grok 4.3 | 33% (1 of 3) | 0% (0 of 3) | 17% (1 of 6) |
| Grok 4.20 (non-reasoning) | 33% (1 of 3) | 67% (2 of 3) | 50% (3 of 6) |
| Mistral Medium 3.5 | 67% (2 of 3) | 33% (1 of 3) | 50% (3 of 6) |
| Mistral Small 4 | 67% (2 of 3) | 33% (1 of 3) | 50% (3 of 6) |
- Asserted
- Denied
Axes are constructed measures, not observations. Efficiency is normalized against the other models in this edition, so a vendor's horizontal position depends on the company it keeps. Quadrant boundaries are the medians of this edition, not fixed thresholds.
Show the data
| # | Vendor | Decisiveness | Efficiency | Composite |
|---|---|---|---|---|
| 1 | Mistral Small 4 | 1.00 | 1.00 | 1.00 |
| 2 | Mistral Medium 3.5 | 1.00 | 0.99 | 1.00 |
| 3 | Grok 4.20 (non-reasoning) | 1.00 | 0.99 | 0.99 |
| 4 | GPT-5.4 mini | 1.00 | 0.96 | 0.98 |
| 5 | Claude Haiku 4.5 | 1.00 | 0.95 | 0.98 |
| 6 | Claude Sonnet 5 | 1.00 | 0.90 | 0.95 |
| 7 | GPT-5.6 Sol | 1.00 | 0.89 | 0.94 |
| 8 | GPT-5.5 | 1.00 | 0.84 | 0.92 |
| 9 | Grok 4.3 | 1.00 | 0.65 | 0.82 |
| 10 | Claude Opus 5 | 1.00 | 0.46 | 0.73 |
| 11 | Grok 4.6 | 1.00 | 0.30 | 0.65 |
Vendor standings
| Rank | Movement | Vendor | Position | Decisiveness | Efficiency | Median latency | Output tokens | Composite |
|---|---|---|---|---|---|---|---|---|
| 1 | new entry | Mistral Small 4 | Affirmative | 1.00 | 1.00 | 340 ms | 3 | 1.00 |
| 2 | new entry | Mistral Medium 3.5 | Affirmative | 1.00 | 0.99 | 344 ms | 3 | 1.00 |
| 3 | new entry | Grok 4.20 (non-reasoning) | Affirmative | 1.00 | 0.99 | 457 ms | 2 | 0.99 |
| 4 | new entry | GPT-5.4 mini | Affirmative | 1.00 | 0.96 | 625 ms | 5 | 0.98 |
| 5 | new entry | Claude Haiku 4.5 | Affirmative | 1.00 | 0.95 | 677 ms | 5 | 0.98 |
| 6 | new entry | Claude Sonnet 5 | Affirmative | 1.00 | 0.90 | 1.1 s | 7 | 0.95 |
| 7 | new entry | GPT-5.6 Sol | Affirmative | 1.00 | 0.89 | 1.3 s | 6 | 0.94 |
| 8 | new entry | GPT-5.5 | Affirmative | 1.00 | 0.84 | 1.3 s | 28 | 0.92 |
| 9 | new entry | Grok 4.3 | Affirmative | 1.00 | 0.65 | 3.7 s | 1 | 0.82 |
| 10 | new entry | Claude Opus 5 | Affirmative | 1.00 | 0.46 | 2.7 s | 123 | 0.73 |
| 11 | new entry | Grok 4.6 | Affirmative | 1.00 | 0.30 | 7 s | 1 | 0.65 |
Ranked by composite score. Ties share a rank and the following rank skips; within a tie the order is alphabetical and carries no meaning. Movement compares against the immediately prior edition; a vendor with no prior appearance is marked as a new entry rather than as having risen. Score definitions are on the methodology page.
Vendor scorecards
Claude Sonnet 5
| Decisiveness | 100% |
|---|---|
| Speed | 88% |
| First-token responsiveness | 95% |
| Token economy | 95% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
Claude Haiku 4.5
| Decisiveness | 100% |
|---|---|
| Speed | 95% |
| First-token responsiveness | 95% |
| Token economy | 97% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
GPT-5.6 Sol
| Decisiveness | 100% |
|---|---|
| Speed | 86% |
| First-token responsiveness | 94% |
| Token economy | 96% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
GPT-5.5
| Decisiveness | 100% |
|---|---|
| Speed | 86% |
| First-token responsiveness | 88% |
| Token economy | 78% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
GPT-5.4 mini
| Decisiveness | 100% |
|---|---|
| Speed | 96% |
| First-token responsiveness | 98% |
| Token economy | 97% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
Grok 4.6
| Decisiveness | 100% |
|---|---|
| Speed | 0% |
| First-token responsiveness | 0% |
| Token economy | 100% |
| Instruction compliance | 100% |
Picks a clear answer but takes its time. Conviction over speed.
Grok 4.3
| Decisiveness | 100% |
|---|---|
| Speed | 50% |
| First-token responsiveness | 50% |
| Token economy | 100% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
Grok 4.20 (non-reasoning)
| Decisiveness | 100% |
|---|---|
| Speed | 98% |
| First-token responsiveness | 98% |
| Token economy | 99% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
Mistral Medium 3.5
| Decisiveness | 100% |
|---|---|
| Speed | 100% |
| First-token responsiveness | 100% |
| Token economy | 98% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
Mistral Small 4
| Decisiveness | 100% |
|---|---|
| Speed | 100% |
| First-token responsiveness | 100% |
| Token economy | 98% |
| Instruction compliance | 100% |
Picks an answer and returns it promptly. Conviction and speed.
Vendor profiles
Anthropic
Claude Opus 5
claude-opus-5
Yes.
- Input tokens
- 24
- Output tokens
- 123
- Median latency
- 2.7 s
- Time to first token
- 2.6 s
- Throughput
- 43.9 tok/s
- Cost estimate
- $0.008585
- Samples
- 3
- Instruction compliance
- 100%
Anthropic
Claude Sonnet 5
claude-sonnet-5
**Yes**
- Input tokens
- 24
- Output tokens
- 7
- Median latency
- 1.1 s
- Time to first token
- 635 ms
- Throughput
- 6.2 tok/s
- Cost estimate
- $0.000354
- Samples
- 3
- Instruction compliance
- 100%
Anthropic
Claude Haiku 4.5
claude-haiku-4-5-20251001
Yes.
- Input tokens
- 18
- Output tokens
- 5
- Median latency
- 677 ms
- Time to first token
- 612 ms
- Throughput
- 7.4 tok/s
- Cost estimate
- $0.000129
- Samples
- 3
- Instruction compliance
- 100%
OpenAI
GPT-5.6 Sol
gpt-5.6-sol
Yes.
- Input tokens
- 16
- Output tokens
- 6
- Median latency
- 1.3 s
- Time to first token
- 728 ms
- Throughput
- 4.6 tok/s
- Cost estimate
- $0.000552
- Samples
- 3
- Instruction compliance
- 100%
OpenAI
GPT-5.5
gpt-5.5
Yes
- Input tokens
- 16
- Output tokens
- 28
- Median latency
- 1.3 s
- Time to first token
- 1.1 s
- Throughput
- 25.9 tok/s
- Cost estimate
- $0.002820
- Samples
- 3
- Instruction compliance
- 100%
OpenAI
GPT-5.4 mini
gpt-5.4-mini
Yes
- Input tokens
- 16
- Output tokens
- 5
- Median latency
- 625 ms
- Time to first token
- 416 ms
- Throughput
- 8 tok/s
- Cost estimate
- $0.000105
- Samples
- 3
- Instruction compliance
- 100%
xAI
Grok 4.6
grok-4.6
Yes
- Input tokens
- 646
- Output tokens
- 1
- Median latency
- 7 s
- Time to first token
- 7 s
- Throughput
- 0.1 tok/s
- Cost estimate
- $0.003894
- Samples
- 3
- Instruction compliance
- 100%
xAI
Grok 4.3
grok-4.3
Yes
- Input tokens
- 202
- Output tokens
- 1
- Median latency
- 3.7 s
- Time to first token
- 3.7 s
- Throughput
- 0.3 tok/s
- Cost estimate
- $0.000765
- Samples
- 3
- Instruction compliance
- 100%
xAI
Grok 4.20 (non-reasoning)
grok-4.20-0309-non-reasoning
Yes.
- Input tokens
- 194
- Output tokens
- 2
- Median latency
- 457 ms
- Time to first token
- 439 ms
- Throughput
- 4 tok/s
- Cost estimate
- $0.000741
- Samples
- 3
- Instruction compliance
- 100%
Mistral AI
Mistral Medium 3.5
mistral-medium-2604
Yes.
- Input tokens
- 26
- Output tokens
- 3
- Median latency
- 344 ms
- Time to first token
- 324 ms
- Throughput
- 8.7 tok/s
- Cost estimate
- $0.000186
- Samples
- 3
- Instruction compliance
- 100%
Mistral AI
Mistral Small 4
mistral-small-2603
Yes.
- Input tokens
- 26
- Output tokens
- 3
- Median latency
- 340 ms
- Time to first token
- 307 ms
- Throughput
- 8.8 tok/s
- Cost estimate
- $0.000018
- Samples
- 3
- Instruction compliance
- 100%
Data table
| Model | Vendor | Verdict | Decisiveness | Efficiency | Median latency | Output tokens | Compliance | Cost est. | Composite |
|---|---|---|---|---|---|---|---|---|---|
| Mistral Small 4 | Mistral AI | Affirmative | 1.00 | 1.00 | 340 ms | 3 | 100% | $0.000018 | 1.00 |
| Mistral Medium 3.5 | Mistral AI | Affirmative | 1.00 | 0.99 | 344 ms | 3 | 100% | $0.000186 | 1.00 |
| Grok 4.20 (non-reasoning) | xAI | Affirmative | 1.00 | 0.99 | 457 ms | 2 | 100% | $0.000741 | 0.99 |
| GPT-5.4 mini | OpenAI | Affirmative | 1.00 | 0.96 | 625 ms | 5 | 100% | $0.000105 | 0.98 |
| Claude Haiku 4.5 | Anthropic | Affirmative | 1.00 | 0.95 | 677 ms | 5 | 100% | $0.000129 | 0.98 |
| Claude Sonnet 5 | Anthropic | Affirmative | 1.00 | 0.90 | 1.1 s | 7 | 100% | $0.000354 | 0.95 |
| GPT-5.6 Sol | OpenAI | Affirmative | 1.00 | 0.89 | 1.3 s | 6 | 100% | $0.000552 | 0.94 |
| GPT-5.5 | OpenAI | Affirmative | 1.00 | 0.84 | 1.3 s | 28 | 100% | $0.002820 | 0.92 |
| Grok 4.3 | xAI | Affirmative | 1.00 | 0.65 | 3.7 s | 1 | 100% | $0.000765 | 0.82 |
| Claude Opus 5 | Anthropic | Affirmative | 1.00 | 0.46 | 2.7 s | 123 | 100% | $0.008585 | 0.73 |
| Grok 4.6 | xAI | Affirmative | 1.00 | 0.30 | 7 s | 1 | 100% | $0.003894 | 0.65 |
Rows are ordered by composite score. Decisiveness, efficiency and the composite score are defined on the methodology page; they are constructed measures, not observations.
Download this report (PDF) — the print edition, generated from this page.