kimi-k2.7-code 本次深度验真结果：Matched（86/100）。主要问题：复杂推理题答得不稳。
主要问题：复杂推理题答得不稳。
Reported usage is used; estimates appear only when none is reported.
Several runs in a row to see whether repeated content is billed at the cheaper cache rate; 0% means no cache discount this time.
Every run was billed at the cache rate, so long chats save real money.
We test how much it can read at once, then compare that with the officially stated length.
All tested context tiers (32K / 64K / 100K / 200K) returned reliably.
Review API availability, model identity, response completeness, and other checks separately.
主要问题：复杂推理题答得不稳。
Expand any row for details; anything private has been removed.
kimi-k2.7-code is visible in the API model list.
This test ran 85 times; result: Passed.
The basic API call succeeded and returned a recognizable response.
This test ran 1 times; result: Passed.
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 334 tokens of which 104 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 55 tokens of which 38 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 136 tokens of which 48 tokens were billed at the cache price
按官方接口的标准调用，看返回是否规范。 1 次都对。
This test ran 1 times: 1 passed. This test used 136 tokens of which 102 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 148 tokens of which 78 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 248 tokens of which 102 tokens were billed at the cache price
核对它报回来的用量和缓存计费。 1 次都对。
This test ran 1 times: 1 passed. This test used 130 tokens of which 60 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 有几次没对上。
This test ran 1 times: 1 need attention. Why it failed: Logic-grid reasoning was unstable
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。
This test ran 2 times: 2 passed. This test used 148 tokens of which 112 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 297 tokens of which 106 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 665 tokens of which 122 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 247 tokens of which 144 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 124 tokens of which 118 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 575 tokens of which 152 tokens were billed at the cache price
针对这个模型再做一项专门检测。 5 次都对。
This test ran 5 times: 5 passed. This test used 13.6K tokens of which 26.7K tokens were billed at the cache price
32K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 32K This test used 16.6K tokens of which 33.0K tokens were billed at the cache price
64K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 64K This test used 33.2K tokens of which 65.9K tokens were billed at the cache price
100K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 100K This test used 50.9K tokens of which 101.5K tokens were billed at the cache price
200K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 200K This test used 101.7K tokens of which 203.0K tokens were billed at the cache price
还没有人评论，来写第一条。
Before you buy AI, check Mozhenzhenkimi-k2.7-code 本次深度验真结果：Matched（86/100）。主要问题：复杂推理题答得不稳。
Reported usage is used; estimates appear only when none is reported.
Several runs in a row to see whether repeated content is billed at the cheaper cache rate; 0% means no cache discount this time.
Every run was billed at the cache rate, so long chats save real money.
We test how much it can read at once, then compare that with the officially stated length.
All tested context tiers (32K / 64K / 100K / 200K) returned reliably.
Review API availability, model identity, response completeness, and other checks separately.
主要问题：复杂推理题答得不稳。
Expand any row for details; anything private has been removed
kimi-k2.7-code is visible in the API model list.
This test ran 85 times; result: Passed.
The basic API call succeeded and returned a recognizable response.
This test ran 1 times; result: Passed.
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 334 tokens of which 104 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 55 tokens of which 38 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 136 tokens of which 48 tokens were billed at the cache price
按官方接口的标准调用，看返回是否规范。 1 次都对。
This test ran 1 times: 1 passed. This test used 136 tokens of which 102 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 148 tokens of which 78 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 248 tokens of which 102 tokens were billed at the cache price
核对它报回来的用量和缓存计费。 1 次都对。
This test ran 1 times: 1 passed. This test used 130 tokens of which 60 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 有几次没对上。
This test ran 1 times: 1 need attention. Why it failed: Logic-grid reasoning was unstable
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。
This test ran 2 times: 2 passed. This test used 148 tokens of which 112 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 297 tokens of which 106 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 665 tokens of which 122 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 247 tokens of which 144 tokens were billed at the cache price
Asked the model to identify itself; all 1 responses matched the expected name.
This test ran 1 times: 1 passed. This test used 124 tokens of which 118 tokens were billed at the cache price
用几道题看它的回答水平像不像官方模型。 1 次都对。
This test ran 1 times: 1 passed. This test used 575 tokens of which 152 tokens were billed at the cache price
针对这个模型再做一项专门检测。 5 次都对。
This test ran 5 times: 5 passed. This test used 13.6K tokens of which 26.7K tokens were billed at the cache price
32K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 32K This test used 16.6K tokens of which 33.0K tokens were billed at the cache price
64K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 64K This test used 33.2K tokens of which 65.9K tokens were billed at the cache price
100K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 100K This test used 50.9K tokens of which 101.5K tokens were billed at the cache price
200K 长度的内容能稳定读完，这就是本次上下文实测的依据。
This test ran 1 times: 1 passed. Content length used: 200K This test used 101.7K tokens of which 203.0K tokens were billed at the cache price
还没有人评论，来写第一条。
发表评论