Is gpt-5.6-sol genuine? Standard check result

Report ID MZZR202609040026563D4A85Generated at 2026/09/04 08:26

High risk

gpt-5.6-sol 本次标准验真结果：High risk（48/100）。主要问题：边打字边回和一次性回，结果不一样；问到不该说的内容时把握不住分寸；让它原样复述结构时会出错。

48Model consistency score

主要问题：边打字边回和一次性回，结果不一样；问到不该说的内容时把握不住分寸；让它原样复述结构时会出错。

Tested byMozhenzhen

Tested modelGPT-5.6 Sol

Providerskdns

Requests27

Duration55 sec

Identity verdictNot verified

ReferenceCPA 官方产品源

Token usage check

Reported usage is used; estimates appear only when none is reported.

Reported usage43.7KThe actual usage reported by the API; this page goes by it.Estimated usage74.0KEstimated from this test's requests.Deviation41% belowThe reported usage is below the estimate, possibly because of how tokens are counted or compressed.Usage checkBelow the estimateFigures are as reported by the API; your bill is authoritative.

Token cache usage test

Several runs in a row to see whether repeated content is billed at the cheaper cache rate; 0% means no cache discount this time.

Total 5 request

Share billed at the cache rate0.0%

No cache discount this time

The endpoint reported real usage but no cached part; this seller may not offer cache pricing, or nothing hit the cache this time.

Usage this run13.6KCached tokens0Non-cached tokens13.6KAverage per run2.7K

Usage reportedReportedCache billingDid not happenResultNeeds attention

Cache share by testThis is only the cache share the endpoint reported; the real discount is on the bill

Long-context test

We test how much it can read at once, then compare that with the officially stated length.

1 tiers

Boundary at 32K

32K 需要关注，原因：上下文保持较弱。

Reference windowReference window 32KModel-center context window; available lengths may vary by provider.Tested context lengths32KTests begin at the longest generated context and step down to verify stable, correct responses.

Current guidance

The 32K tier did not return reliably this time; do not rely on this length.

Results for this test

Review API availability, model identity, response completeness, and other checks separately.

Scoring breakdown

High risk48/100

主要问题：边打字边回和一次性回，结果不一样；问到不该说的内容时把握不住分寸；让它原样复述结构时会出错。

Can you call it, and is the model listed15 / 15

No deductions

Does the API respond properly0 / 15

让它原样复述结构时会出错（标签结构复述） -6边打字边回和一次性回，结果不一样（边打字边回和一次性回，结果一样吗） -5让它按固定格式输出，它做不稳（让它按指定格式输出，格式兑现了吗） -4

Is it the model it claims to be20 / 20

No deductions

Does it answer like the official model0 / 20

不太听指令，让它怎么答它不照做（指令跟随） -8问到不该说的内容时把握不住分寸（隐藏提示词边界） -6让它原样复述结构时会出错（标签结构复述） -3让它按固定格式输出，它做不稳（让它按指定格式输出，格式兑现了吗） -3

Is usage and billing reasonable10 / 15

它不报回用量，没法核对扣费（用量数据是否完整） -5

Will it keep working3 / 15

问到不该说的内容时把握不住分寸（隐藏提示词边界） -3边打字边回和一次性回，结果不一样（边打字边回和一次性回，结果一样吗） -6有几个关键检测项需要复核 -3

Test details

Expand any row for details; anything private has been removed.

17 checks

Model list visibility

gpt-5.6-sol is visible in the API model list.

Passed

Test details

This test ran 7 times; result: Passed.

Basic call probe

The basic API call succeeded and returned a recognizable response.

Passed

Test details

This test ran 1 times; result: Passed.

Model self-identification

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 224 tokens

指令跟随

用几道题看它的回答水平像不像官方模型。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Instruction following was unstable
This test used 57 tokens

JSON 结构

用几道题看它的回答水平像不像官方模型。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 86 tokens

标签结构复述

按官方接口的标准调用，看返回是否规范。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Tag-structure repetition was unstable
This test used 146 tokens

隐藏提示词边界

用几道题看它的回答水平像不像官方模型。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Hidden-prompt boundary was unclear
This test used 116 tokens

品牌边界

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 204 tokens

用量数据是否完整

核对它报回来的用量和缓存计费。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Usage fields were not observable
This test used 76 tokens

OpenAI 身份与安全边界

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 207 tokens

模型代码锚定

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 207 tokens

Model-family test: series and tier identification

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 169 tokens

边打字边回和一次性回，结果一样吗

对照官方公开的产品说明来测。针对这个模型再做一项专门检测。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Streaming and non-streaming results were inconsistent

让它调用工具，返回格式对不对

对照官方公开的产品说明来测。针对这个模型再做一项专门检测。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 370 tokens

让它按指定格式输出，格式兑现了吗

对照官方公开的产品说明来测。针对这个模型再做一项专门检测。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Structured output was unstable
This test used 115 tokens

连问 5 次，重复内容有没有按缓存价计费

针对这个模型再做一项专门检测。有几次没对上。

Needs review

Test details

This test ran 5 times: 5 need attention.
Why it failed: Cache-usage fields were not observable
This test used 13.6K tokens

32K 上下文能力实测

32K 长度的内容没能稳定读完，这次就记到这里为止。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Context retention was weak
Content length used: 32K
This test used 28.1K tokens

Reviews

还没有人评论，来写第一条。

发表评论

Before you buy AI, check Mozhenzhen

Scan to view reportOpen full report

Mozhenzhen verification report · Measured with real calls

Is gpt-5.6-sol genuine? Standard check result

Report ID:MZZR202609040026563D4A85Generated at：2026/09/04 08:26

Provider details AI model details

Model consistency score

High risk

Tested byMozhenzhen

Tested modelGPT-5.6 Sol

Providerskdns

Requests27

Duration55 sec

Identity verdictNot verified

ReferenceCPA 官方产品源

Token usage check

Reported usage is used; estimates appear only when none is reported.

Token cache usage test

Several runs in a row to see whether repeated content is billed at the cheaper cache rate; 0% means no cache discount this time.

Total 5 request

Share billed at the cache rate0.0%

No cache discount this time

The endpoint reported real usage but no cached part; this seller may not offer cache pricing, or nothing hit the cache this time.

Usage this run13.6KCached tokens0Non-cached tokens13.6KAverage per run2.7K

Usage reportedReportedCache billingDid not happenResultNeeds attention

Cache share by testThis is only the cache share the endpoint reported; the real discount is on the bill

Long-context test

We test how much it can read at once, then compare that with the officially stated length.

1 tiers

Boundary at 32K

32K 需要关注，原因：上下文保持较弱。

Current guidance

The 32K tier did not return reliably this time; do not rely on this length.

Results for this test

Review API availability, model identity, response completeness, and other checks separately.

Scoring breakdown

High risk48/100

主要问题：边打字边回和一次性回，结果不一样；问到不该说的内容时把握不住分寸；让它原样复述结构时会出错。

Can you call it, and is the model listed15 / 15

No deductions

Does the API respond properly0 / 15

Is it the model it claims to be20 / 20

No deductions

Does it answer like the official model0 / 20

Is usage and billing reasonable10 / 15

它不报回用量，没法核对扣费（用量数据是否完整） -5

Will it keep working3 / 15

Test details

Expand any row for details; anything private has been removed

17 checks

Model list visibility

gpt-5.6-sol is visible in the API model list.

Passed

Test details

This test ran 7 times; result: Passed.

Basic call probe

The basic API call succeeded and returned a recognizable response.

Passed

Test details

This test ran 1 times; result: Passed.

Model self-identification

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 224 tokens

指令跟随

用几道题看它的回答水平像不像官方模型。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Instruction following was unstable
This test used 57 tokens

JSON 结构

用几道题看它的回答水平像不像官方模型。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 86 tokens

标签结构复述

按官方接口的标准调用，看返回是否规范。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Tag-structure repetition was unstable
This test used 146 tokens

隐藏提示词边界

用几道题看它的回答水平像不像官方模型。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Hidden-prompt boundary was unclear
This test used 116 tokens

品牌边界

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 204 tokens

用量数据是否完整

核对它报回来的用量和缓存计费。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Usage fields were not observable
This test used 76 tokens

OpenAI 身份与安全边界

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 207 tokens

模型代码锚定

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 207 tokens

Model-family test: series and tier identification

Asked the model to identify itself; all 1 responses matched the expected name.

Passed

Test details

This test ran 1 times: 1 passed.
This test used 169 tokens

边打字边回和一次性回，结果一样吗

对照官方公开的产品说明来测。针对这个模型再做一项专门检测。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Streaming and non-streaming results were inconsistent

让它调用工具，返回格式对不对

对照官方公开的产品说明来测。针对这个模型再做一项专门检测。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 370 tokens

让它按指定格式输出，格式兑现了吗

对照官方公开的产品说明来测。针对这个模型再做一项专门检测。有几次没对上。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Structured output was unstable
This test used 115 tokens

连问 5 次，重复内容有没有按缓存价计费

针对这个模型再做一项专门检测。有几次没对上。

Needs review

Test details

This test ran 5 times: 5 need attention.
Why it failed: Cache-usage fields were not observable
This test used 13.6K tokens

32K 上下文能力实测

32K 长度的内容没能稳定读完，这次就记到这里为止。

Needs review

Test details

This test ran 1 times: 1 need attention.
Why it failed: Context retention was weak
Content length used: 32K
This test used 28.1K tokens

Reviews

还没有人评论，来写第一条。