pi-hypercharm-provider 1.3.18 → 1.3.20

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (3) hide show
  1. package/README.md +6 -5
  2. package/models.json +137 -95
  3. package/package.json +1 -1
package/README.md CHANGED
@@ -15,7 +15,7 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
15
15
 
16
16
  ## Features
17
17
 
18
- - **33+ AI Models** including DeepSeek V4 Flash/Pro, GLM 5/5.1, Kimi K2.5/K2.6, Qwen3.6/3.7, MiniMax M2.7, Gemma 4, GPT-OSS, and Llama
18
+ - **34+ AI Models** including DeepSeek V4 Flash/Pro, GLM 5/5.1, Kimi K2.5/K2.6, Qwen3.6/3.7, MiniMax M2.7, Gemma 4, GPT-OSS, and Llama
19
19
  - **DeepSeek Native Thinking** — Uses the `deepseek` thinking format for Charm Hyper requests, with native `reasoning_effort` on models that publish levels
20
20
  - **OpenAI-compatible API** via Charm Hyper's `/v1/chat/completions` endpoint
21
21
  - **OAuth Device Flow** — sign in with `/login` under the `hypercharm` provider (independent of the official provider's `hyper` registration)
@@ -31,11 +31,12 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
31
31
  | DeepSeek V4 Flash 0731 | Text | 1.0M | 384K | $0.44 | $1.32 |
32
32
  | DeepSeek V4 Pro | Text | 1.0M | 384K | $2.40 | $4.80 |
33
33
  | DeepSeek V4 Pro 0813 | Text | 1.0M | 262K | $1.44 | $4.31 |
34
- | Gemma 4 26B A4B | Text | 256K | 26K | $0.11 | $0.37 |
35
- | GLM 5.3 | Text | 1.0M | 262K | $1.52 | $4.79 |
34
+ | DeepSeek V4.1 Flash | Text + Image | 1.0M | 26K | $0.30 | $1.20 |
35
+ | Gemma 4 26B A4B | Text | 256K | 26K | $0.12 | $0.42 |
36
+ | GLM 5.3 | Text + Image | 1.0M | 262K | $1.52 | $4.79 |
36
37
  | GLM 5.3 Flash | Text + Image | 1.0M | 131K | $0.16 | $0.54 |
37
38
  | GLM-5 | Text | 203K | 20K | $0.86 | $2.78 |
38
- | GLM-5.1 | Text | 203K | 3K | $1.36 | $4.27 |
39
+ | GLM-5.1 | Text | 203K | 3K | $1.33 | $4.22 |
39
40
  | GLM-5.2 | Text | 1.0M | 33K | $1.52 | $4.79 |
40
41
  | gpt-oss-120b | Text | 128K | 13K | $0.18 | $0.68 |
41
42
  | Inkling | Text + Image | 1.0M | 33K | $1.09 | $4.41 |
@@ -46,7 +47,7 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
46
47
  | Kimi K3 | Text + Image | 1.0M | 16K | $3.27 | $16.33 |
47
48
  | Llama 3.3 70B Instruct | Text | 128K | 13K | $0.61 | $1.04 |
48
49
  | Llama 4 Maverick 17B 128E Instruct FP8 | Text | 430K | 43K | $0.27 | $0.90 |
49
- | MiniMax M2.7 | Text | 262K | 7K | $0.46 | $1.71 |
50
+ | MiniMax M2.7 | Text | 262K | 7K | $0.40 | $1.50 |
50
51
  | MiniMax M3 | Text + Image | 512K | 512K | $0.33 | $1.31 |
51
52
  | Qwen3 Coder 480B A35B Instruct INT4 Mixed AR | Text | 106K | 11K | $0.45 | $2.15 |
52
53
  | Qwen3 Next 80B A3B Instruct | Text | 262K | 26K | $0.12 | $1.14 |
package/models.json CHANGED
@@ -35,7 +35,7 @@
35
35
  "name": "DeepSeek V4 Flash 0731",
36
36
  "reasoning": true,
37
37
  "thinkingLevelMap": {
38
- "off": "none",
38
+ "off": null,
39
39
  "minimal": null,
40
40
  "low": "low",
41
41
  "medium": null,
@@ -123,6 +123,38 @@
123
123
  "maxTokensField": "max_tokens"
124
124
  }
125
125
  },
126
+ {
127
+ "id": "deepseek-v4.1-flash",
128
+ "name": "DeepSeek V4.1 Flash",
129
+ "reasoning": true,
130
+ "thinkingLevelMap": {
131
+ "off": null,
132
+ "minimal": null,
133
+ "low": "low",
134
+ "medium": null,
135
+ "high": "high",
136
+ "xhigh": "xhigh",
137
+ "max": null
138
+ },
139
+ "input": [
140
+ "text",
141
+ "image"
142
+ ],
143
+ "cost": {
144
+ "input": 0.3,
145
+ "output": 1.2,
146
+ "cacheRead": 0.03,
147
+ "cacheWrite": 0
148
+ },
149
+ "contextWindow": 1048576,
150
+ "maxTokens": 26214,
151
+ "compat": {
152
+ "supportsStore": false,
153
+ "supportsReasoningEffort": true,
154
+ "thinkingFormat": "deepseek",
155
+ "maxTokensField": "max_tokens"
156
+ }
157
+ },
126
158
  {
127
159
  "id": "gemma-4-26b-a4b-it",
128
160
  "name": "Gemma 4 26B A4B",
@@ -140,10 +172,10 @@
140
172
  "text"
141
173
  ],
142
174
  "cost": {
143
- "input": 0.106,
144
- "output": 0.368,
145
- "cacheRead": 0,
146
- "cacheWrite": 0.053
175
+ "input": 0.122,
176
+ "output": 0.42,
177
+ "cacheRead": 0.061,
178
+ "cacheWrite": 0
147
179
  },
148
180
  "contextWindow": 256000,
149
181
  "maxTokens": 25600,
@@ -168,7 +200,8 @@
168
200
  "max": "max"
169
201
  },
170
202
  "input": [
171
- "text"
203
+ "text",
204
+ "image"
172
205
  ],
173
206
  "cost": {
174
207
  "input": 1.52432,
@@ -236,8 +269,8 @@
236
269
  "cost": {
237
270
  "input": 0.86,
238
271
  "output": 2.784,
239
- "cacheRead": 0,
240
- "cacheWrite": 0.43
272
+ "cacheRead": 0.43,
273
+ "cacheWrite": 0
241
274
  },
242
275
  "contextWindow": 202752,
243
276
  "maxTokens": 20275,
@@ -265,10 +298,10 @@
265
298
  "text"
266
299
  ],
267
300
  "cost": {
268
- "input": 1.358,
269
- "output": 4.268,
270
- "cacheRead": 0,
271
- "cacheWrite": 0.679
301
+ "input": 1.326,
302
+ "output": 4.22,
303
+ "cacheRead": 0.663,
304
+ "cacheWrite": 0
272
305
  },
273
306
  "contextWindow": 202750,
274
307
  "maxTokens": 3276,
@@ -329,8 +362,8 @@
329
362
  "cost": {
330
363
  "input": 0.178,
331
364
  "output": 0.68,
332
- "cacheRead": 0,
333
- "cacheWrite": 0.089
365
+ "cacheRead": 0.089,
366
+ "cacheWrite": 0
334
367
  },
335
368
  "contextWindow": 128072,
336
369
  "maxTokens": 13107,
@@ -392,8 +425,8 @@
392
425
  "cost": {
393
426
  "input": 0.6,
394
427
  "output": 2.5,
395
- "cacheRead": 0,
396
- "cacheWrite": 0.3
428
+ "cacheRead": 0.3,
429
+ "cacheWrite": 0
397
430
  },
398
431
  "contextWindow": 262144,
399
432
  "maxTokens": 26214,
@@ -423,8 +456,8 @@
423
456
  "cost": {
424
457
  "input": 0.5584,
425
458
  "output": 2.935,
426
- "cacheRead": 0,
427
- "cacheWrite": 0.2792
459
+ "cacheRead": 0.2792,
460
+ "cacheWrite": 0
428
461
  },
429
462
  "contextWindow": 262144,
430
463
  "maxTokens": 26214,
@@ -550,8 +583,8 @@
550
583
  "cost": {
551
584
  "input": 0.6066,
552
585
  "output": 1.0386,
553
- "cacheRead": 0,
554
- "cacheWrite": 0.3033
586
+ "cacheRead": 0.3033,
587
+ "cacheWrite": 0
555
588
  },
556
589
  "contextWindow": 128000,
557
590
  "maxTokens": 12800,
@@ -581,8 +614,8 @@
581
614
  "cost": {
582
615
  "input": 0.274,
583
616
  "output": 0.8992,
584
- "cacheRead": 0,
585
- "cacheWrite": 0.137
617
+ "cacheRead": 0.137,
618
+ "cacheWrite": 0
586
619
  },
587
620
  "contextWindow": 430000,
588
621
  "maxTokens": 43000,
@@ -601,10 +634,10 @@
601
634
  "text"
602
635
  ],
603
636
  "cost": {
604
- "input": 0.458,
605
- "output": 1.712,
606
- "cacheRead": 0,
607
- "cacheWrite": 0.229
637
+ "input": 0.404,
638
+ "output": 1.496,
639
+ "cacheRead": 0.202,
640
+ "cacheWrite": 0
608
641
  },
609
642
  "contextWindow": 262100,
610
643
  "maxTokens": 6553,
@@ -657,8 +690,8 @@
657
690
  "cost": {
658
691
  "input": 0.445,
659
692
  "output": 2.145,
660
- "cacheRead": 0,
661
- "cacheWrite": 0.2225
693
+ "cacheRead": 0.2225,
694
+ "cacheWrite": 0
662
695
  },
663
696
  "contextWindow": 106000,
664
697
  "maxTokens": 10600,
@@ -679,8 +712,8 @@
679
712
  "cost": {
680
713
  "input": 0.1175,
681
714
  "output": 1.136,
682
- "cacheRead": 0,
683
- "cacheWrite": 0.05875
715
+ "cacheRead": 0.05875,
716
+ "cacheWrite": 0
684
717
  },
685
718
  "contextWindow": 262144,
686
719
  "maxTokens": 26214,
@@ -696,13 +729,13 @@
696
729
  "name": "Qwen3.6-Flash",
697
730
  "reasoning": true,
698
731
  "thinkingLevelMap": {
699
- "off": "off",
700
- "minimal": null,
701
- "low": null,
702
- "medium": null,
703
- "high": null,
732
+ "off": "none",
733
+ "minimal": "minimal",
734
+ "low": "low",
735
+ "medium": "medium",
736
+ "high": "high",
704
737
  "xhigh": null,
705
- "max": "max"
738
+ "max": null
706
739
  },
707
740
  "input": [
708
741
  "text",
@@ -718,7 +751,7 @@
718
751
  "maxTokens": 64000,
719
752
  "compat": {
720
753
  "supportsStore": false,
721
- "supportsReasoningEffort": false,
754
+ "supportsReasoningEffort": true,
722
755
  "thinkingFormat": "deepseek",
723
756
  "maxTokensField": "max_tokens"
724
757
  }
@@ -728,13 +761,13 @@
728
761
  "name": "Qwen3.6-Max",
729
762
  "reasoning": true,
730
763
  "thinkingLevelMap": {
731
- "off": "off",
732
- "minimal": null,
733
- "low": null,
734
- "medium": null,
735
- "high": null,
764
+ "off": "none",
765
+ "minimal": "minimal",
766
+ "low": "low",
767
+ "medium": "medium",
768
+ "high": "high",
736
769
  "xhigh": null,
737
- "max": "max"
770
+ "max": null
738
771
  },
739
772
  "input": [
740
773
  "text"
@@ -749,7 +782,7 @@
749
782
  "maxTokens": 64000,
750
783
  "compat": {
751
784
  "supportsStore": false,
752
- "supportsReasoningEffort": false,
785
+ "supportsReasoningEffort": true,
753
786
  "thinkingFormat": "deepseek",
754
787
  "maxTokensField": "max_tokens"
755
788
  }
@@ -759,13 +792,13 @@
759
792
  "name": "Qwen3.6-Plus",
760
793
  "reasoning": true,
761
794
  "thinkingLevelMap": {
762
- "off": "off",
763
- "minimal": null,
764
- "low": null,
765
- "medium": null,
766
- "high": null,
795
+ "off": "none",
796
+ "minimal": "minimal",
797
+ "low": "low",
798
+ "medium": "medium",
799
+ "high": "high",
767
800
  "xhigh": null,
768
- "max": "max"
801
+ "max": null
769
802
  },
770
803
  "input": [
771
804
  "text",
@@ -781,7 +814,7 @@
781
814
  "maxTokens": 64000,
782
815
  "compat": {
783
816
  "supportsStore": false,
784
- "supportsReasoningEffort": false,
817
+ "supportsReasoningEffort": true,
785
818
  "thinkingFormat": "deepseek",
786
819
  "maxTokensField": "max_tokens"
787
820
  }
@@ -791,13 +824,13 @@
791
824
  "name": "Qwen3.7-Flash",
792
825
  "reasoning": true,
793
826
  "thinkingLevelMap": {
794
- "off": "off",
795
- "minimal": null,
796
- "low": null,
797
- "medium": null,
798
- "high": null,
827
+ "off": "none",
828
+ "minimal": "minimal",
829
+ "low": "low",
830
+ "medium": "medium",
831
+ "high": "high",
799
832
  "xhigh": null,
800
- "max": "max"
833
+ "max": null
801
834
  },
802
835
  "input": [
803
836
  "text",
@@ -813,7 +846,7 @@
813
846
  "maxTokens": 64000,
814
847
  "compat": {
815
848
  "supportsStore": false,
816
- "supportsReasoningEffort": false,
849
+ "supportsReasoningEffort": true,
817
850
  "thinkingFormat": "deepseek",
818
851
  "maxTokensField": "max_tokens"
819
852
  }
@@ -823,13 +856,13 @@
823
856
  "name": "Qwen3.7-Max",
824
857
  "reasoning": true,
825
858
  "thinkingLevelMap": {
826
- "off": "off",
827
- "minimal": null,
828
- "low": null,
829
- "medium": null,
830
- "high": null,
859
+ "off": "none",
860
+ "minimal": "minimal",
861
+ "low": "low",
862
+ "medium": "medium",
863
+ "high": "high",
831
864
  "xhigh": null,
832
- "max": "max"
865
+ "max": null
833
866
  },
834
867
  "input": [
835
868
  "text"
@@ -844,7 +877,7 @@
844
877
  "maxTokens": 64000,
845
878
  "compat": {
846
879
  "supportsStore": false,
847
- "supportsReasoningEffort": false,
880
+ "supportsReasoningEffort": true,
848
881
  "thinkingFormat": "deepseek",
849
882
  "maxTokensField": "max_tokens"
850
883
  }
@@ -854,13 +887,13 @@
854
887
  "name": "Qwen3.7-Plus",
855
888
  "reasoning": true,
856
889
  "thinkingLevelMap": {
857
- "off": "off",
858
- "minimal": null,
859
- "low": null,
860
- "medium": null,
861
- "high": null,
890
+ "off": "none",
891
+ "minimal": "minimal",
892
+ "low": "low",
893
+ "medium": "medium",
894
+ "high": "high",
862
895
  "xhigh": null,
863
- "max": "max"
896
+ "max": null
864
897
  },
865
898
  "input": [
866
899
  "text",
@@ -876,7 +909,7 @@
876
909
  "maxTokens": 64000,
877
910
  "compat": {
878
911
  "supportsStore": false,
879
- "supportsReasoningEffort": false,
912
+ "supportsReasoningEffort": true,
880
913
  "thinkingFormat": "deepseek",
881
914
  "maxTokensField": "max_tokens"
882
915
  }
@@ -884,7 +917,16 @@
884
917
  {
885
918
  "id": "qwen3.8-2.4t-a95b",
886
919
  "name": "Qwen3.8-2.4T-A95B",
887
- "reasoning": false,
920
+ "reasoning": true,
921
+ "thinkingLevelMap": {
922
+ "off": null,
923
+ "minimal": "minimal",
924
+ "low": "low",
925
+ "medium": "medium",
926
+ "high": "high",
927
+ "xhigh": null,
928
+ "max": null
929
+ },
888
930
  "input": [
889
931
  "text"
890
932
  ],
@@ -898,7 +940,7 @@
898
940
  "maxTokens": 128000,
899
941
  "compat": {
900
942
  "supportsStore": false,
901
- "supportsReasoningEffort": false,
943
+ "supportsReasoningEffort": true,
902
944
  "thinkingFormat": "deepseek",
903
945
  "maxTokensField": "max_tokens"
904
946
  }
@@ -908,13 +950,13 @@
908
950
  "name": "Qwen3.8-27B",
909
951
  "reasoning": true,
910
952
  "thinkingLevelMap": {
911
- "off": "off",
912
- "minimal": null,
913
- "low": null,
914
- "medium": null,
915
- "high": null,
953
+ "off": "none",
954
+ "minimal": "minimal",
955
+ "low": "low",
956
+ "medium": "medium",
957
+ "high": "high",
916
958
  "xhigh": null,
917
- "max": "max"
959
+ "max": null
918
960
  },
919
961
  "input": [
920
962
  "text",
@@ -930,7 +972,7 @@
930
972
  "maxTokens": 128000,
931
973
  "compat": {
932
974
  "supportsStore": false,
933
- "supportsReasoningEffort": false,
975
+ "supportsReasoningEffort": true,
934
976
  "thinkingFormat": "deepseek",
935
977
  "maxTokensField": "max_tokens"
936
978
  }
@@ -940,13 +982,13 @@
940
982
  "name": "Qwen3.8-Flash",
941
983
  "reasoning": true,
942
984
  "thinkingLevelMap": {
943
- "off": "off",
944
- "minimal": null,
945
- "low": null,
946
- "medium": null,
947
- "high": null,
985
+ "off": "none",
986
+ "minimal": "minimal",
987
+ "low": "low",
988
+ "medium": "medium",
989
+ "high": "high",
948
990
  "xhigh": null,
949
- "max": "max"
991
+ "max": null
950
992
  },
951
993
  "input": [
952
994
  "text",
@@ -962,7 +1004,7 @@
962
1004
  "maxTokens": 128000,
963
1005
  "compat": {
964
1006
  "supportsStore": false,
965
- "supportsReasoningEffort": false,
1007
+ "supportsReasoningEffort": true,
966
1008
  "thinkingFormat": "deepseek",
967
1009
  "maxTokensField": "max_tokens"
968
1010
  }
@@ -972,13 +1014,13 @@
972
1014
  "name": "Qwen3.8-Max",
973
1015
  "reasoning": true,
974
1016
  "thinkingLevelMap": {
975
- "off": "off",
976
- "minimal": null,
977
- "low": null,
978
- "medium": null,
979
- "high": null,
1017
+ "off": "none",
1018
+ "minimal": "minimal",
1019
+ "low": "low",
1020
+ "medium": "medium",
1021
+ "high": "high",
980
1022
  "xhigh": null,
981
- "max": "max"
1023
+ "max": null
982
1024
  },
983
1025
  "input": [
984
1026
  "text",
@@ -994,7 +1036,7 @@
994
1036
  "maxTokens": 65536,
995
1037
  "compat": {
996
1038
  "supportsStore": false,
997
- "supportsReasoningEffort": false,
1039
+ "supportsReasoningEffort": true,
998
1040
  "thinkingFormat": "deepseek",
999
1041
  "maxTokensField": "max_tokens"
1000
1042
  }
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "pi-hypercharm-provider",
3
- "version": "1.3.18",
3
+ "version": "1.3.20",
4
4
  "description": "HyperCharm provider extension for pi - Access DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, and GPT-OSS models through the Charm Hyper API",
5
5
  "type": "module",
6
6
  "main": "index.ts",