pi-hypercharm-provider 1.3.4 → 1.3.6

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 (4) hide show
  1. package/README.md +13 -10
  2. package/models.json +141 -47
  3. package/package.json +3 -3
  4. package/pnpm-workspace.yaml +7 -7
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
- - **26+ 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
+ - **29+ 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
  - **Official Catalog Sync** from Charm's typed `/v1/provider` endpoint, matching `@charmland/pi-hyper-provider`
@@ -27,21 +27,21 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
27
27
  | Model | Type | Context | Max Tokens | Input Cost | Output Cost |
28
28
  |-------|------|---------|------------|------------|-------------|
29
29
  | DeepSeek V4 Flash | Text | 1.0M | 384K | $0.20 | $0.40 |
30
- | DeepSeek V4 Flash 0731 | Text | 1.0M | 384K | $0.48 | $1.44 |
30
+ | DeepSeek V4 Flash 0731 | Text | 1.0M | 384K | $0.44 | $1.32 |
31
31
  | DeepSeek V4 Pro | Text | 1.0M | 384K | $2.40 | $4.80 |
32
32
  | DeepSeek V4 Pro 0813 | Text | 1.0M | 262K | $1.44 | $4.31 |
33
33
  | Gemma 4 26B A4B | Text | 256K | 26K | $0.11 | $0.41 |
34
- | GLM-5 | Text | 203K | 20K | $0.91 | $2.93 |
35
- | GLM-5.1 | Text | 203K | 3K | $1.36 | $4.40 |
36
- | GLM-5.2 | Text + Image | 1.0M | 33K | $1.52 | $4.79 |
37
- | gpt-oss-120b | Text | 131K | 13K | $0.19 | $0.70 |
34
+ | GLM-5 | Text | 203K | 20K | $0.90 | $2.80 |
35
+ | GLM-5.1 | Text | 203K | 3K | $1.31 | $4.27 |
36
+ | GLM-5.2 | Text | 1.0M | 33K | $1.52 | $4.79 |
37
+ | gpt-oss-120b | Text | 128K | 13K | $0.18 | $0.68 |
38
38
  | Kimi K2.5 | Text | 262K | 26K | $0.54 | $2.85 |
39
39
  | Kimi K2.6 | Text + Image | 262K | 26K | $1.03 | $4.36 |
40
- | Kimi K2.7 Code | Text + Image | 256K | 16K | $1.03 | $4.36 |
40
+ | Kimi K2.7 Code | Text + Image | 262K | 16K | $1.03 | $4.36 |
41
41
  | Kimi K3 | Text + Image | 1.0M | 16K | $3.27 | $16.33 |
42
- | Llama 3.3 70B Instruct | Text | 128K | 13K | $0.64 | $0.77 |
43
- | Llama 4 Maverick 17B 128E Instruct FP8 | Text | 430K | 43K | $0.27 | $0.90 |
44
- | MiniMax M2.7 | Text | 262K | 7K | $0.41 | $1.51 |
42
+ | Llama 3.3 70B Instruct | Text | 128K | 13K | $0.61 | $1.04 |
43
+ | Llama 4 Maverick 17B 128E Instruct FP8 | Text | 430K | 43K | $0.28 | $0.93 |
44
+ | MiniMax M2.7 | Text | 262K | 7K | $0.40 | $1.50 |
45
45
  | MiniMax M3 | Text + Image | 512K | 512K | $0.33 | $1.31 |
46
46
  | Qwen3 Coder 480B A35B Instruct INT4 Mixed AR | Text | 106K | 11K | $0.45 | $2.15 |
47
47
  | Qwen3 Next 80B A3B Instruct | Text | 262K | 26K | $0.12 | $1.14 |
@@ -51,6 +51,9 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
51
51
  | Qwen3.7-Flash | Text + Image | 1.0M | 64K | $0.20 | $0.80 |
52
52
  | Qwen3.7-Max | Text | 1.0M | 64K | $2.50 | $7.50 |
53
53
  | Qwen3.7-Plus | Text + Image | 1.0M | 64K | $1.20 | $4.80 |
54
+ | Qwen3.8-2.4T-A95B | Text | 1.0M | 128K | $2.00 | $6.00 |
55
+ | Qwen3.8-27B | Text + Image | 1.0M | 128K | $0.50 | $3.00 |
56
+ | Qwen3.8-Flash | Text + Image | 1.0M | 128K | $0.15 | $0.47 |
54
57
  | Qwen3.8-Max | Text + Image | 1.0M | 66K | $2.00 | $6.00 |
55
58
  *Costs are per million tokens. Prices subject to change — check [hyper.charm.land](https://hyper.charm.land) for current pricing.*
56
59
 
package/models.json CHANGED
@@ -36,23 +36,23 @@
36
36
  "reasoning": true,
37
37
  "thinkingLevelMap": {
38
38
  "off": "none",
39
- "minimal": "minimal",
39
+ "minimal": null,
40
40
  "low": "low",
41
- "medium": "medium",
41
+ "medium": null,
42
42
  "high": "high",
43
- "xhigh": "xhigh",
43
+ "xhigh": null,
44
44
  "max": "max"
45
45
  },
46
46
  "input": [
47
47
  "text"
48
48
  ],
49
49
  "cost": {
50
- "input": 0.479072,
51
- "output": 1.437216,
52
- "cacheRead": 0.0152432,
50
+ "input": 0.44,
51
+ "output": 1.32,
52
+ "cacheRead": 0.044,
53
53
  "cacheWrite": 0
54
54
  },
55
- "contextWindow": 1048576,
55
+ "contextWindow": 1000000,
56
56
  "maxTokens": 384000,
57
57
  "compat": {
58
58
  "supportsStore": false,
@@ -99,11 +99,11 @@
99
99
  "thinkingLevelMap": {
100
100
  "off": "none",
101
101
  "minimal": null,
102
- "low": null,
102
+ "low": "low",
103
103
  "medium": null,
104
104
  "high": "high",
105
- "xhigh": "xhigh",
106
- "max": null
105
+ "xhigh": null,
106
+ "max": "max"
107
107
  },
108
108
  "input": [
109
109
  "text"
@@ -171,10 +171,10 @@
171
171
  "text"
172
172
  ],
173
173
  "cost": {
174
- "input": 0.91,
175
- "output": 2.934,
174
+ "input": 0.9,
175
+ "output": 2.804,
176
176
  "cacheRead": 0,
177
- "cacheWrite": 0.455
177
+ "cacheWrite": 0.45
178
178
  },
179
179
  "contextWindow": 202752,
180
180
  "maxTokens": 20275,
@@ -202,10 +202,10 @@
202
202
  "text"
203
203
  ],
204
204
  "cost": {
205
- "input": 1.36,
206
- "output": 4.4,
205
+ "input": 1.314,
206
+ "output": 4.268,
207
207
  "cacheRead": 0,
208
- "cacheWrite": 0.68
208
+ "cacheWrite": 0.657
209
209
  },
210
210
  "contextWindow": 202750,
211
211
  "maxTokens": 3276,
@@ -221,17 +221,16 @@
221
221
  "name": "GLM-5.2",
222
222
  "reasoning": true,
223
223
  "thinkingLevelMap": {
224
- "off": "none",
224
+ "off": null,
225
225
  "minimal": null,
226
226
  "low": null,
227
227
  "medium": null,
228
228
  "high": "high",
229
- "xhigh": null,
230
- "max": "max"
229
+ "xhigh": "xhigh",
230
+ "max": null
231
231
  },
232
232
  "input": [
233
- "text",
234
- "image"
233
+ "text"
235
234
  ],
236
235
  "cost": {
237
236
  "input": 1.52432,
@@ -239,7 +238,7 @@
239
238
  "cacheRead": 0.152432,
240
239
  "cacheWrite": 0
241
240
  },
242
- "contextWindow": 1048576,
241
+ "contextWindow": 1000000,
243
242
  "maxTokens": 32768,
244
243
  "compat": {
245
244
  "supportsStore": false,
@@ -253,24 +252,24 @@
253
252
  "name": "gpt-oss-120b",
254
253
  "reasoning": true,
255
254
  "thinkingLevelMap": {
256
- "off": null,
257
- "minimal": null,
255
+ "off": "none",
256
+ "minimal": "minimal",
258
257
  "low": "low",
259
258
  "medium": "medium",
260
259
  "high": "high",
261
- "xhigh": null,
262
- "max": null
260
+ "xhigh": "xhigh",
261
+ "max": "max"
263
262
  },
264
263
  "input": [
265
264
  "text"
266
265
  ],
267
266
  "cost": {
268
- "input": 0.188,
269
- "output": 0.7,
267
+ "input": 0.178,
268
+ "output": 0.68,
270
269
  "cacheRead": 0,
271
- "cacheWrite": 0.094
270
+ "cacheWrite": 0.089
272
271
  },
273
- "contextWindow": 131072,
272
+ "contextWindow": 128072,
274
273
  "maxTokens": 13107,
275
274
  "compat": {
276
275
  "supportsStore": false,
@@ -315,13 +314,13 @@
315
314
  "name": "Kimi K2.6",
316
315
  "reasoning": true,
317
316
  "thinkingLevelMap": {
318
- "off": "off",
317
+ "off": null,
319
318
  "minimal": null,
320
- "low": null,
321
- "medium": null,
322
- "high": null,
319
+ "low": "low",
320
+ "medium": "medium",
321
+ "high": "high",
323
322
  "xhigh": null,
324
- "max": "max"
323
+ "max": null
325
324
  },
326
325
  "input": [
327
326
  "text",
@@ -337,7 +336,7 @@
337
336
  "maxTokens": 26214,
338
337
  "compat": {
339
338
  "supportsStore": false,
340
- "supportsReasoningEffort": false,
339
+ "supportsReasoningEffort": true,
341
340
  "thinkingFormat": "deepseek",
342
341
  "maxTokensField": "max_tokens"
343
342
  }
@@ -345,7 +344,16 @@
345
344
  {
346
345
  "id": "kimi-k2.7-code",
347
346
  "name": "Kimi K2.7 Code",
348
- "reasoning": false,
347
+ "reasoning": true,
348
+ "thinkingLevelMap": {
349
+ "off": "off",
350
+ "minimal": null,
351
+ "low": null,
352
+ "medium": null,
353
+ "high": null,
354
+ "xhigh": null,
355
+ "max": "max"
356
+ },
349
357
  "input": [
350
358
  "text",
351
359
  "image"
@@ -356,7 +364,7 @@
356
364
  "cacheRead": 0.206872,
357
365
  "cacheWrite": 0
358
366
  },
359
- "contextWindow": 256000,
367
+ "contextWindow": 262000,
360
368
  "maxTokens": 16000,
361
369
  "compat": {
362
370
  "supportsStore": false,
@@ -414,10 +422,10 @@
414
422
  "text"
415
423
  ],
416
424
  "cost": {
417
- "input": 0.638,
418
- "output": 0.768,
425
+ "input": 0.6066,
426
+ "output": 1.0386,
419
427
  "cacheRead": 0,
420
- "cacheWrite": 0.319
428
+ "cacheWrite": 0.3033
421
429
  },
422
430
  "contextWindow": 128000,
423
431
  "maxTokens": 12800,
@@ -445,10 +453,10 @@
445
453
  "text"
446
454
  ],
447
455
  "cost": {
448
- "input": 0.274,
449
- "output": 0.8992,
456
+ "input": 0.284,
457
+ "output": 0.934,
450
458
  "cacheRead": 0,
451
- "cacheWrite": 0.137
459
+ "cacheWrite": 0.142
452
460
  },
453
461
  "contextWindow": 430000,
454
462
  "maxTokens": 43000,
@@ -467,10 +475,10 @@
467
475
  "text"
468
476
  ],
469
477
  "cost": {
470
- "input": 0.408,
471
- "output": 1.512,
478
+ "input": 0.404,
479
+ "output": 1.496,
472
480
  "cacheRead": 0,
473
- "cacheWrite": 0.204
481
+ "cacheWrite": 0.202
474
482
  },
475
483
  "contextWindow": 262100,
476
484
  "maxTokens": 6553,
@@ -747,6 +755,92 @@
747
755
  "maxTokensField": "max_tokens"
748
756
  }
749
757
  },
758
+ {
759
+ "id": "qwen3.8-2.4t-a95b",
760
+ "name": "Qwen3.8-2.4T-A95B",
761
+ "reasoning": false,
762
+ "input": [
763
+ "text"
764
+ ],
765
+ "cost": {
766
+ "input": 2,
767
+ "output": 6,
768
+ "cacheRead": 0.25,
769
+ "cacheWrite": 0
770
+ },
771
+ "contextWindow": 1000000,
772
+ "maxTokens": 128000,
773
+ "compat": {
774
+ "supportsStore": false,
775
+ "supportsReasoningEffort": false,
776
+ "thinkingFormat": "deepseek",
777
+ "maxTokensField": "max_tokens"
778
+ }
779
+ },
780
+ {
781
+ "id": "qwen3.8-27b",
782
+ "name": "Qwen3.8-27B",
783
+ "reasoning": true,
784
+ "thinkingLevelMap": {
785
+ "off": "off",
786
+ "minimal": null,
787
+ "low": null,
788
+ "medium": null,
789
+ "high": null,
790
+ "xhigh": null,
791
+ "max": "max"
792
+ },
793
+ "input": [
794
+ "text",
795
+ "image"
796
+ ],
797
+ "cost": {
798
+ "input": 0.5,
799
+ "output": 3,
800
+ "cacheRead": 0.1,
801
+ "cacheWrite": 0
802
+ },
803
+ "contextWindow": 1000000,
804
+ "maxTokens": 128000,
805
+ "compat": {
806
+ "supportsStore": false,
807
+ "supportsReasoningEffort": false,
808
+ "thinkingFormat": "deepseek",
809
+ "maxTokensField": "max_tokens"
810
+ }
811
+ },
812
+ {
813
+ "id": "qwen3.8-flash",
814
+ "name": "Qwen3.8-Flash",
815
+ "reasoning": true,
816
+ "thinkingLevelMap": {
817
+ "off": "off",
818
+ "minimal": null,
819
+ "low": null,
820
+ "medium": null,
821
+ "high": null,
822
+ "xhigh": null,
823
+ "max": "max"
824
+ },
825
+ "input": [
826
+ "text",
827
+ "image"
828
+ ],
829
+ "cost": {
830
+ "input": 0.15,
831
+ "output": 0.47,
832
+ "cacheRead": 0.016,
833
+ "cacheWrite": 0
834
+ },
835
+ "contextWindow": 1000000,
836
+ "maxTokens": 128000,
837
+ "compat": {
838
+ "supportsStore": false,
839
+ "supportsReasoningEffort": false,
840
+ "thinkingFormat": "deepseek",
841
+ "maxTokensField": "max_tokens"
842
+ }
843
+ },
750
844
  {
751
845
  "id": "qwen3.8-max",
752
846
  "name": "Qwen3.8-Max",
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "pi-hypercharm-provider",
3
- "version": "1.3.4",
3
+ "version": "1.3.6",
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",
@@ -22,8 +22,8 @@
22
22
  "gpt-oss"
23
23
  ],
24
24
  "devDependencies": {
25
- "@earendil-works/pi-ai": "0.84.3",
26
- "@earendil-works/pi-coding-agent": "0.84.3",
25
+ "@earendil-works/pi-ai": "0.84.4",
26
+ "@earendil-works/pi-coding-agent": "0.84.4",
27
27
  "@types/node": "^24.10.1",
28
28
  "typescript": "6.0.3"
29
29
  },
package/pnpm-workspace.yaml CHANGED
@@ -6,10 +6,10 @@ onlyBuiltDependencies:
6
6
  - protobufjs
7
7
 
8
8
  minimumReleaseAgeExclude:
9
- - '@earendil-works/pi-agent-core@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
10
- - '@earendil-works/pi-ai@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
11
- - '@earendil-works/pi-client@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
12
- - '@earendil-works/pi-coding-agent@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
13
- - '@earendil-works/pi-protocol@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
14
- - '@earendil-works/pi-telemetry@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
15
- - '@earendil-works/pi-tui@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3'
9
+ - '@earendil-works/pi-agent-core@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'
10
+ - '@earendil-works/pi-ai@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'
11
+ - '@earendil-works/pi-client@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'
12
+ - '@earendil-works/pi-coding-agent@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'
13
+ - '@earendil-works/pi-protocol@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'
14
+ - '@earendil-works/pi-telemetry@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'
15
+ - '@earendil-works/pi-tui@0.84.0 || 0.84.1 || 0.84.2 || 0.84.3 || 0.84.4'