[Bug] Model Catalog Aggregation. #22

New issue

Closed

opened 2026-08-28 09:08:39 +00:00 by epistemophiliac · 0 comments

epistemophiliac commented

2026-08-28 09:08:39 +00:00

Owner

1	
LongCat 2.0
3 providers
200/hr
1M ctx
Tools

100
19
76–99
—	
0.872
best of 3

569	
meituan/longcat-2.0:free
NR-1-chat
Tools

no data
no data
53
—	0.527

	#	Model	
Reliability	
Speed	
Intelligence	
Guardrails
Score
On
5	
tencent/hy3
3 providers
20 RPM
131K ctx
Tools

98
38
76
—	
0.788
best of 3

65	
Tencent Hy3
2 providers
$0.10/mo credit
262K ctx
Tools

40
60
99
—	
0.736
best of 2

87	
tencent/hy3:free
NR-1-chat
Tools

87
65
53
—	0.682	

144	
Tencent HY 3
kilo
free
262K ctx
Tools

no data
no data
76
—	0.631

We aggregate the paid endpoints of the model names, but not the free models as well for that model? There some aggregation missing, and Its not even using NR-1-chat at all? Even though, longchat-2.0 is our 1st model rn. Its using alot, on other providers. Is it much quicker on NR-1-chat, which will Improve our usage on longchat-2.0. There is also other models that are messed up, and not fully aggregated together as well. That are missing some, we need to Improve our regex that causes and calculates the aggregation.

``` 1 LongCat 2.0 3 providers 200/hr 1M ctx Tools 100 19 76–99 — 0.872 best of 3 569 meituan/longcat-2.0:free NR-1-chat Tools no data no data 53 — 0.527 ``` ``` # Model Reliability Speed Intelligence Guardrails Score On 5 tencent/hy3 3 providers 20 RPM 131K ctx Tools 98 38 76 — 0.788 best of 3 65 Tencent Hy3 2 providers $0.10/mo credit 262K ctx Tools 40 60 99 — 0.736 best of 2 87 tencent/hy3:free NR-1-chat Tools 87 65 53 — 0.682 144 Tencent HY 3 kilo free 262K ctx Tools no data no data 76 — 0.631 ``` We aggregate the paid endpoints of the model names, but not the free models as well for that model? There some aggregation missing, and Its not even using NR-1-chat at all? Even though, longchat-2.0 is our 1st model rn. Its using alot, on other providers. Is it much quicker on NR-1-chat, which will Improve our usage on longchat-2.0. There is also other models that are messed up, and not fully aggregated together as well. That are missing some, we need to Improve our regex that causes and calculates the aggregation.

epistemophiliac referenced this issue from a commit

2026-08-28 10:42:15 +00:00

fix(upstream-sync): target develop instead of main; aggregate :free model_ids (#22)

epistemophiliac referenced this issue from a pull request that will close it,

2026-08-28 10:46:40 +00:00

fix(upstream-sync): target develop instead of main; aggregate :free model_ids (#22) #24

epistemophiliac referenced this issue from a commit

2026-08-28 12:49:49 +00:00

fix(model-groups): strip leading brand names, normalize number spacing (#22)

epistemophiliac closed this issue

2026-08-28 14:46:46 +00:00

epistemophiliac referenced this issue from a commit