pi-hypercharm-provider 1.3.19 → 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 +7 -6
  2. package/models.json +64 -31
  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)
@@ -31,13 +31,14 @@ _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 | $1.39 | $4.36 |
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
- | gpt-oss-120b | Text | 128K | 13K | $0.18 | $0.61 |
41
+ | gpt-oss-120b | Text | 128K | 13K | $0.18 | $0.68 |
41
42
  | Inkling | Text + Image | 1.0M | 33K | $1.09 | $4.41 |
42
43
  | Kimi K2 Thinking | Text | 262K | 26K | $0.60 | $2.50 |
43
44
  | Kimi K2.5 | Text | 262K | 26K | $0.56 | $2.94 |
@@ -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
@@ -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": 1.386,
269
- "output": 4.356,
270
- "cacheRead": 0,
271
- "cacheWrite": 0.693
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,
@@ -327,10 +360,10 @@
327
360
  "text"
328
361
  ],
329
362
  "cost": {
330
- "input": 0.18,
331
- "output": 0.61,
332
- "cacheRead": 0,
333
- "cacheWrite": 0.09
363
+ "input": 0.178,
364
+ "output": 0.68,
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,
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "pi-hypercharm-provider",
3
- "version": "1.3.19",
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",