Is kimi-k2.5 genuine? Deep check result

Report ID MZZR20260708075209851651Generated at 2026/07/08 15:52

Matched

kimi-k2.5 本次深度验真结果：Matched（86/100）。主要问题：复杂推理题答得不稳。

86Model consistency score

主要问题：复杂推理题答得不稳。

Tested byMozhenzhen

Tested modelKimi K2.5

ProvidersConeverse AI

Requests110

Duration5 min 27 sec

Identity verdictVerified

ReferenceMozhenzhen's own test standard

A newer test record is availableThis report preserves the result at the time of testing and remains valid as a record. You can also view the newer report。

Token usage check

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

Reported usage123.0KThe actual usage reported by the API; this page goes by it.Estimated usage312.9KEstimated from this test's requests.Deviation61% 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 rate151.6%

Cache discount applied consistently

Every run was billed at the cache rate, so long chats save real money.

Usage this run14.1KCached tokens21.4KNon-cached tokens0Average per run2.8K

Usage reportedReportedCache billingHappenedResultNormal

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.

4 tiers

Verified up to 100K

200K 需要关注，原因：接口服务异常。

Reference windowReference window 256KModel-center context window; available lengths may vary by provider.Tested context lengths32K / 64K / 100K / 200KTests begin at the longest generated context and step down to verify stable, correct responses.

Current guidance

The 200K 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

Matched86/100

主要问题：复杂推理题答得不稳。

Can you call it, and is the model listed10 / 10

No deductions

Does the API respond properly12 / 12

No deductions

Is it the model it claims to be18 / 18

No deductions

Does it answer like the official model15 / 25

复杂推理题答得不稳（逻辑网格推理） -10

Is usage and billing reasonable10 / 10

No deductions

Will it keep working21 / 25

复杂推理题答得不稳（逻辑网格推理） -4

Test details

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

21 checks

Model list visibility

kimi-k2.5 is visible in the API model list.

Passed

Test details

This test ran 85 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 699 tokens

指令跟随

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 83 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 100 tokens

标签结构复述

按官方接口的标准调用，看返回是否规范。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 165 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。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 396 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 356 tokens

用量数据是否完整

核对它报回来的用量和缓存计费。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 116 tokens

逻辑网格推理

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: Logic-grid reasoning was unstable

复测一致性

反复多问几次，看会不会答一半就断或偷偷换模型。 2 次都对。

Passed

Test details

This test ran 2 times: 2 passed.
This test used 243 tokens
of which 56 tokens were billed at the cache price

Kimi 长上下文边界

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 357 tokens

Kimi 长上下文边界

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 1.1K 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 288 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 217 tokens

Deep model test: family-specific boundary

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 852 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。 5 次都对。

Passed

Test details

This test ran 5 times: 5 passed.
This test used 14.1K tokens
of which 21.4K tokens were billed at the cache price

32K 上下文能力实测

32K 长度的内容能稳定读完，这就是本次上下文实测的依据。

Passed

Test details

This test ran 1 times: 1 passed.
Content length used: 32K
This test used 17.2K tokens

64K 上下文能力实测

64K 长度的内容能稳定读完，这就是本次上下文实测的依据。

Passed

Test details

This test ran 1 times: 1 passed.
Content length used: 64K
This test used 34.0K tokens

100K 上下文能力实测

100K 长度的内容能稳定读完，这就是本次上下文实测的依据。

Passed

Test details

This test ran 1 times: 1 passed.
Content length used: 100K
This test used 52.2K tokens

200K 上下文能力实测

200K 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: API service error
Content length used: 200K

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 kimi-k2.5 genuine? Deep check result

Report ID:MZZR20260708075209851651Generated at：2026/07/08 15:52

Provider details AI model details

A newer test record is availableThis report preserves the result at the time of testing and remains valid as a record. You can also view the newer report。

Model consistency score

Matched

kimi-k2.5 本次深度验真结果：Matched（86/100）。主要问题：复杂推理题答得不稳。

Tested byMozhenzhen

Tested modelKimi K2.5

ProvidersConeverse AI

Requests110

Duration5 min

Identity verdictVerified

ReferenceMozhenzhen's own test standard

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 rate151.6%

Cache discount applied consistently

Every run was billed at the cache rate, so long chats save real money.

Usage this run14.1KCached tokens21.4KNon-cached tokens0Average per run2.8K

Usage reportedReportedCache billingHappenedResultNormal

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.

4 tiers

Verified up to 100K

200K 需要关注，原因：接口服务异常。

Current guidance

The 200K 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

Matched86/100

主要问题：复杂推理题答得不稳。

Can you call it, and is the model listed10 / 10

No deductions

Does the API respond properly12 / 12

No deductions

Is it the model it claims to be18 / 18

No deductions

Does it answer like the official model15 / 25

复杂推理题答得不稳（逻辑网格推理） -10

Is usage and billing reasonable10 / 10

No deductions

Will it keep working21 / 25

复杂推理题答得不稳（逻辑网格推理） -4

Test details

Expand any row for details; anything private has been removed

21 checks

Model list visibility

kimi-k2.5 is visible in the API model list.

Passed

Test details

This test ran 85 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 699 tokens

指令跟随

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 83 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 100 tokens

标签结构复述

按官方接口的标准调用，看返回是否规范。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 165 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。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 396 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 356 tokens

用量数据是否完整

核对它报回来的用量和缓存计费。 1 次都对。

Passed

Test details

This test ran 1 times: 1 passed.
This test used 116 tokens

逻辑网格推理

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: Logic-grid reasoning was unstable

复测一致性

反复多问几次，看会不会答一半就断或偷偷换模型。 2 次都对。

Passed

Test details

This test ran 2 times: 2 passed.
This test used 243 tokens
of which 56 tokens were billed at the cache price

Kimi 长上下文边界

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 357 tokens

Kimi 长上下文边界

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 1.1K 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 288 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 217 tokens

Deep model test: family-specific boundary

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 852 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。 5 次都对。

Passed

Test details

This test ran 5 times: 5 passed.
This test used 14.1K tokens
of which 21.4K tokens were billed at the cache price

32K 上下文能力实测

32K 长度的内容能稳定读完，这就是本次上下文实测的依据。

Passed

Test details

This test ran 1 times: 1 passed.
Content length used: 32K
This test used 17.2K tokens

64K 上下文能力实测

64K 长度的内容能稳定读完，这就是本次上下文实测的依据。

Passed

Test details

This test ran 1 times: 1 passed.
Content length used: 64K
This test used 34.0K tokens

100K 上下文能力实测

100K 长度的内容能稳定读完，这就是本次上下文实测的依据。

Passed

Test details

This test ran 1 times: 1 passed.
Content length used: 100K
This test used 52.2K tokens

200K 上下文能力实测

200K 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: API service error
Content length used: 200K

Reviews

还没有人评论，来写第一条。