pi-fireworks-provider 1.1.1 → 1.3.0

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 +14 -10
  2. package/package.json +11 -12
  3. package/patch.json +42 -119
package/README.md CHANGED
@@ -15,7 +15,7 @@ _Kimi, MiniMax, GLM, DeepSeek, GPT-OSS — via Fireworks AI's Anthropic Messages
15
15
 
16
16
  ## Features
17
17
 
18
- - **35+ AI Models** including Kimi K2.5, MiniMax M2.5, GLM 4.5/4.7/5, DeepSeek V3.1/V3.2, DeepSeek V4 Flash, and GPT-OSS
18
+ - **39+ AI Models** including Kimi K2.5, MiniMax M2.5, GLM 4.5/4.7/5, DeepSeek V3.1/V3.2, DeepSeek V4 Flash, and GPT-OSS
19
19
  - **Dual API support** via Fireworks AI's Anthropic Messages and OpenAI-compatible completions endpoints (per-model routing, matching pi core's Fireworks provider)
20
20
  - **Service tiers** — toggle Fireworks `priority` vs `standard` per request on supported models (with priority pricing reflected in cost tracking), via a keybinding, `/fireworks-tier`, and a footer status area
21
21
  - **Preserved thinking** — toggle Fireworks' `reasoning_history: "preserved"` so prior assistant reasoning is retained across turns (better multi-turn recall; uses more tokens), via the `/fireworks-settings` panel, with a model-select notification. Matches neuralwatt/makora's settings-only UX, adapted to Fireworks' single global `reasoning_history` knob
@@ -73,8 +73,8 @@ pi
73
73
  |-------|------|---------|------------|------------|-------------|
74
74
  | DeepSeek V3.1 | Text | 164K | 164K | $0.56 | $1.68 |
75
75
  | DeepSeek V3.2 | Text | 164K | 160K | $0.56 | $1.68 |
76
- | DeepSeek V4 Flash | Text | 1.0M | 1.0M | $0.14 | $0.28 |
77
- | DeepSeek V4 Pro | Text | 1.0M | 1.0M | Free | Free |
76
+ | DeepSeek V4 Flash | Text | 1.0M | 384K | $0.14 | $0.28 |
77
+ | DeepSeek V4 Pro | Text | 1.0M | 384K | $1.74 | $3.48 |
78
78
  | DeepSeek V4 Pro (router) | Text | 1.0M | 1.0M | $1.74 | $3.48 |
79
79
  | Gemma 4 26B A4B IT | Text + Image | 262K | 0 | Free | Free |
80
80
  | Gemma 4 31B IT | Text + Image | 262K | 0 | Free | Free |
@@ -82,27 +82,31 @@ pi
82
82
  | GLM 4.5 Air | Text | 131K | 131K | $0.22 | $0.88 |
83
83
  | GLM 4.7 | Text | 203K | 198K | $0.60 | $2.20 |
84
84
  | GLM 5 | Text | 203K | 131K | $1.00 | $3.20 |
85
- | GLM 5 Fast (router) | Text | 203K | 131K | $1.00 | $3.20 |
85
+ | GLM 5 Fast | Text | 203K | 131K | $1.00 | $3.20 |
86
86
  | GLM 5.1 | Text | 203K | 131K | $1.40 | $4.40 |
87
- | GLM 5.1 Fast (router) | Text | 203K | 131K | $1.40 | $4.40 |
88
- | GLM 5.2 | Text | 1.0M | 0 | Free | Free |
87
+ | GLM 5.1 Fast | Text | 203K | 131K | $2.80 | $8.80 |
88
+ | GLM 5.2 | Text | 1.0M | 131K | $1.40 | $4.40 |
89
+ | GLM 5.2 Fast | Text | 1.0M | 131K | $2.10 | $6.60 |
89
90
  | GPT OSS 120B | Text | 131K | 33K | $0.15 | $0.60 |
90
- | GPT OSS 20B | Text | 131K | 33K | $0.05 | $0.20 |
91
+ | GPT OSS 20B | Text | 131K | 33K | $0.07 | $0.30 |
91
92
  | Kimi K2 Instruct | Text | 131K | 16K | $1.00 | $3.00 |
92
93
  | Kimi K2 Thinking | Text | 262K | 256K | $0.60 | $2.50 |
93
94
  | Kimi K2.5 | Text + Image | 262K | 256K | $0.60 | $3.00 |
94
95
  | Kimi K2.5 Fast (router) | Text + Image | 262K | 256K | $0.60 | $3.00 |
95
96
  | Kimi K2.6 | Text + Image | 262K | 262K | $0.95 | $4.00 |
96
97
  | Kimi K2.6 (router) | Text + Image | 262K | 262K | $0.95 | $4.00 |
97
- | Kimi K2.6 Turbo (router) | Text + Image | 262K | 262K | $0.95 | $4.00 |
98
- | Kimi K2.7 Code | Text + Image | 262K | 0 | Free | Free |
98
+ | Kimi K2.6 Fast | Text + Image | 262K | 262K | $2.00 | $8.00 |
99
+ | Kimi K2.6 Turbo | Text + Image | 262K | 262K | $2.00 | $8.00 |
100
+ | Kimi K2.7 Code | Text + Image | 262K | 262K | $0.95 | $4.00 |
101
+ | Kimi K2.7 Code Fast | Text + Image | 262K | 262K | $1.90 | $8.00 |
99
102
  | Llama 3.3 70B Instruct | Text | 131K | 0 | Free | Free |
100
103
  | MiniMax M2.7 (router) | Text | 204K | 0 | $0.30 | $1.20 |
101
- | Minimax M3 | Text + Image | 512K | 0 | Free | Free |
102
104
  | MiniMax-M2.1 | Text | 197K | 200K | $0.30 | $1.20 |
103
105
  | MiniMax-M2.5 | Text | 197K | 197K | $0.30 | $1.20 |
104
106
  | MiniMax-M2.7 | Text | 197K | 197K | $0.30 | $1.20 |
107
+ | MiniMax-M3 | Text | 512K | 512K | $0.30 | $1.20 |
105
108
  | NVIDIA Nemotron 3 Ultra NVFP4 | Text | 262K | 0 | Free | Free |
109
+ | Qwen 3.7 Plus | Text + Image | 262K | 66K | $0.40 | $1.60 |
106
110
  | Qwen3 8B | Text | 41K | 0 | Free | Free |
107
111
  | Qwen3 VL 30B A3B Instruct | Text + Image | 262K | 0 | Free | Free |
108
112
  | Qwen3 VL 30B A3B Thinking | Text + Image | 262K | 0 | Free | Free |
package/package.json CHANGED
@@ -1,18 +1,9 @@
1
1
  {
2
2
  "name": "pi-fireworks-provider",
3
- "version": "1.1.1",
3
+ "version": "1.3.0",
4
4
  "description": "Fireworks AI provider extension for pi - Access Kimi, MiniMax, GLM, DeepSeek, and GPT-OSS models through the Fireworks AI API",
5
5
  "type": "module",
6
6
  "main": "index.ts",
7
- "scripts": {
8
- "clean": "echo 'nothing to clean'",
9
- "build": "echo 'nothing to build'",
10
- "check": "tsc --noEmit && vitest run && knip --no-gitignore",
11
- "lint:dead": "knip --no-gitignore",
12
- "test": "vitest run",
13
- "test:watch": "vitest",
14
- "update-models": "node scripts/update-models.js"
15
- },
16
7
  "keywords": [
17
8
  "pi",
18
9
  "extension",
@@ -47,5 +38,13 @@
47
38
  "./index.ts"
48
39
  ]
49
40
  },
50
- "packageManager": "pnpm@11.6.0"
51
- }
41
+ "scripts": {
42
+ "clean": "echo 'nothing to clean'",
43
+ "build": "echo 'nothing to build'",
44
+ "check": "tsc --noEmit && vitest run && knip --no-gitignore",
45
+ "lint:dead": "knip --no-gitignore",
46
+ "test": "vitest run",
47
+ "test:watch": "vitest",
48
+ "update-models": "node scripts/update-models.js"
49
+ }
50
+ }
package/patch.json CHANGED
@@ -33,9 +33,14 @@
33
33
  },
34
34
  "accounts/fireworks/models/deepseek-v4-flash": {
35
35
  "name": "DeepSeek V4 Flash",
36
- "api": "anthropic-messages",
37
- "baseUrl": "https://api.fireworks.ai/inference",
38
36
  "reasoning": true,
37
+ "thinkingLevelMap": {
38
+ "off": "none",
39
+ "minimal": null,
40
+ "low": "high",
41
+ "medium": "high",
42
+ "xhigh": "max"
43
+ },
39
44
  "input": ["text"],
40
45
  "cost": {
41
46
  "input": 0.14,
@@ -46,17 +51,19 @@
46
51
  "contextWindow": 1000000,
47
52
  "maxTokens": 384000,
48
53
  "compat": {
49
- "sendSessionAffinityHeaders": true,
50
- "supportsEagerToolInputStreaming": false,
51
- "supportsCacheControlOnTools": false,
52
- "supportsLongCacheRetention": false
54
+ "supportsReasoningEffort": true
53
55
  }
54
56
  },
55
57
  "accounts/fireworks/models/deepseek-v4-pro": {
56
58
  "name": "DeepSeek V4 Pro",
57
- "api": "anthropic-messages",
58
- "baseUrl": "https://api.fireworks.ai/inference",
59
59
  "reasoning": true,
60
+ "thinkingLevelMap": {
61
+ "off": "none",
62
+ "minimal": null,
63
+ "low": "high",
64
+ "medium": "high",
65
+ "xhigh": "max"
66
+ },
60
67
  "input": ["text"],
61
68
  "cost": {
62
69
  "input": 1.74,
@@ -67,10 +74,7 @@
67
74
  "contextWindow": 1000000,
68
75
  "maxTokens": 384000,
69
76
  "compat": {
70
- "sendSessionAffinityHeaders": true,
71
- "supportsEagerToolInputStreaming": false,
72
- "supportsCacheControlOnTools": false,
73
- "supportsLongCacheRetention": false
77
+ "supportsReasoningEffort": true
74
78
  }
75
79
  },
76
80
  "accounts/fireworks/models/glm-4p5": {
@@ -139,8 +143,6 @@
139
143
  },
140
144
  "accounts/fireworks/models/glm-5p1": {
141
145
  "name": "GLM 5.1",
142
- "api": "anthropic-messages",
143
- "baseUrl": "https://api.fireworks.ai/inference",
144
146
  "reasoning": true,
145
147
  "input": ["text"],
146
148
  "cost": {
@@ -152,10 +154,7 @@
152
154
  "contextWindow": 202800,
153
155
  "maxTokens": 131072,
154
156
  "compat": {
155
- "sendSessionAffinityHeaders": true,
156
- "supportsEagerToolInputStreaming": false,
157
- "supportsCacheControlOnTools": false,
158
- "supportsLongCacheRetention": false
157
+ "supportsReasoningEffort": true
159
158
  }
160
159
  },
161
160
  "accounts/fireworks/models/glm-5p2": {
@@ -186,8 +185,6 @@
186
185
  },
187
186
  "accounts/fireworks/models/gpt-oss-120b": {
188
187
  "name": "GPT OSS 120B",
189
- "api": "anthropic-messages",
190
- "baseUrl": "https://api.fireworks.ai/inference",
191
188
  "reasoning": true,
192
189
  "input": ["text"],
193
190
  "cost": {
@@ -199,16 +196,11 @@
199
196
  "contextWindow": 131072,
200
197
  "maxTokens": 32768,
201
198
  "compat": {
202
- "sendSessionAffinityHeaders": true,
203
- "supportsEagerToolInputStreaming": false,
204
- "supportsCacheControlOnTools": false,
205
- "supportsLongCacheRetention": false
199
+ "supportsReasoningEffort": true
206
200
  }
207
201
  },
208
202
  "accounts/fireworks/models/gpt-oss-20b": {
209
203
  "name": "GPT OSS 20B",
210
- "api": "anthropic-messages",
211
- "baseUrl": "https://api.fireworks.ai/inference",
212
204
  "reasoning": true,
213
205
  "input": ["text"],
214
206
  "cost": {
@@ -220,10 +212,7 @@
220
212
  "contextWindow": 131072,
221
213
  "maxTokens": 32768,
222
214
  "compat": {
223
- "sendSessionAffinityHeaders": true,
224
- "supportsEagerToolInputStreaming": false,
225
- "supportsCacheControlOnTools": false,
226
- "supportsLongCacheRetention": false
215
+ "supportsReasoningEffort": true
227
216
  }
228
217
  },
229
218
  "accounts/fireworks/models/gemma-4-26b-a4b-it": {
@@ -303,8 +292,6 @@
303
292
  },
304
293
  "accounts/fireworks/models/kimi-k2p6": {
305
294
  "name": "Kimi K2.6",
306
- "api": "anthropic-messages",
307
- "baseUrl": "https://api.fireworks.ai/inference",
308
295
  "reasoning": true,
309
296
  "input": ["text", "image"],
310
297
  "cost": {
@@ -316,16 +303,11 @@
316
303
  "contextWindow": 262000,
317
304
  "maxTokens": 262000,
318
305
  "compat": {
319
- "sendSessionAffinityHeaders": true,
320
- "supportsEagerToolInputStreaming": false,
321
- "supportsCacheControlOnTools": false,
322
- "supportsLongCacheRetention": false
306
+ "supportsReasoningEffort": true
323
307
  }
324
308
  },
325
309
  "accounts/fireworks/models/kimi-k2p7-code": {
326
310
  "name": "Kimi K2.7 Code",
327
- "api": "anthropic-messages",
328
- "baseUrl": "https://api.fireworks.ai/inference",
329
311
  "reasoning": true,
330
312
  "input": ["text", "image"],
331
313
  "cost": {
@@ -337,10 +319,7 @@
337
319
  "contextWindow": 262000,
338
320
  "maxTokens": 262000,
339
321
  "compat": {
340
- "sendSessionAffinityHeaders": true,
341
- "supportsEagerToolInputStreaming": false,
342
- "supportsCacheControlOnTools": false,
343
- "supportsLongCacheRetention": false
322
+ "supportsReasoningEffort": true
344
323
  }
345
324
  },
346
325
  "accounts/fireworks/models/minimax-m2p1": {
@@ -377,8 +356,6 @@
377
356
  },
378
357
  "accounts/fireworks/models/minimax-m2p7": {
379
358
  "name": "MiniMax-M2.7",
380
- "api": "anthropic-messages",
381
- "baseUrl": "https://api.fireworks.ai/inference",
382
359
  "reasoning": true,
383
360
  "input": ["text"],
384
361
  "cost": {
@@ -390,16 +367,11 @@
390
367
  "contextWindow": 196608,
391
368
  "maxTokens": 196608,
392
369
  "compat": {
393
- "sendSessionAffinityHeaders": true,
394
- "supportsEagerToolInputStreaming": false,
395
- "supportsCacheControlOnTools": false,
396
- "supportsLongCacheRetention": false
370
+ "supportsReasoningEffort": true
397
371
  }
398
372
  },
399
373
  "accounts/fireworks/models/minimax-m3": {
400
374
  "name": "MiniMax-M3",
401
- "api": "anthropic-messages",
402
- "baseUrl": "https://api.fireworks.ai/inference",
403
375
  "reasoning": true,
404
376
  "input": ["text"],
405
377
  "cost": {
@@ -410,33 +382,12 @@
410
382
  },
411
383
  "contextWindow": 512000,
412
384
  "maxTokens": 512000,
413
- "compat": {
414
- "sendSessionAffinityHeaders": true,
415
- "supportsEagerToolInputStreaming": false,
416
- "supportsCacheControlOnTools": false,
417
- "supportsLongCacheRetention": false
418
- }
419
- },
420
- "accounts/fireworks/models/qwen3p6-plus": {
421
- "name": "Qwen 3.6 Plus",
422
- "reasoning": true,
423
- "input": ["text", "image"],
424
- "cost": {
425
- "input": 0.5,
426
- "output": 3,
427
- "cacheRead": 0.1,
428
- "cacheWrite": 0
429
- },
430
- "contextWindow": 262144,
431
- "maxTokens": 8192,
432
385
  "compat": {
433
386
  "supportsReasoningEffort": true
434
387
  }
435
388
  },
436
389
  "accounts/fireworks/models/qwen3p7-plus": {
437
390
  "name": "Qwen 3.7 Plus",
438
- "api": "anthropic-messages",
439
- "baseUrl": "https://api.fireworks.ai/inference",
440
391
  "reasoning": true,
441
392
  "input": ["text", "image"],
442
393
  "cost": {
@@ -448,15 +399,19 @@
448
399
  "contextWindow": 262144,
449
400
  "maxTokens": 65536,
450
401
  "compat": {
451
- "sendSessionAffinityHeaders": true,
452
- "supportsEagerToolInputStreaming": false,
453
- "supportsCacheControlOnTools": false,
454
- "supportsLongCacheRetention": false
402
+ "supportsReasoningEffort": true
455
403
  }
456
404
  },
457
405
  "accounts/fireworks/routers/deepseek-v4-pro": {
458
406
  "name": "DeepSeek V4 Pro (router)",
459
407
  "reasoning": true,
408
+ "thinkingLevelMap": {
409
+ "off": "none",
410
+ "minimal": null,
411
+ "low": "high",
412
+ "medium": "high",
413
+ "xhigh": "max"
414
+ },
460
415
  "input": ["text"],
461
416
  "cost": {
462
417
  "input": 1.74,
@@ -488,8 +443,6 @@
488
443
  },
489
444
  "accounts/fireworks/routers/glm-5p1-fast": {
490
445
  "name": "GLM 5.1 Fast",
491
- "api": "anthropic-messages",
492
- "baseUrl": "https://api.fireworks.ai/inference",
493
446
  "reasoning": true,
494
447
  "input": ["text"],
495
448
  "cost": {
@@ -501,17 +454,19 @@
501
454
  "contextWindow": 202800,
502
455
  "maxTokens": 131072,
503
456
  "compat": {
504
- "sendSessionAffinityHeaders": true,
505
- "supportsEagerToolInputStreaming": false,
506
- "supportsCacheControlOnTools": false,
507
- "supportsLongCacheRetention": false
457
+ "supportsReasoningEffort": true
508
458
  }
509
459
  },
510
460
  "accounts/fireworks/routers/glm-5p2-fast": {
511
461
  "name": "GLM 5.2 Fast",
512
- "api": "anthropic-messages",
513
- "baseUrl": "https://api.fireworks.ai/inference",
514
462
  "reasoning": true,
463
+ "thinkingLevelMap": {
464
+ "off": "none",
465
+ "minimal": null,
466
+ "low": "high",
467
+ "medium": "high",
468
+ "xhigh": "max"
469
+ },
515
470
  "input": ["text"],
516
471
  "cost": {
517
472
  "input": 2.1,
@@ -522,10 +477,7 @@
522
477
  "contextWindow": 1048575,
523
478
  "maxTokens": 131072,
524
479
  "compat": {
525
- "sendSessionAffinityHeaders": true,
526
- "supportsEagerToolInputStreaming": false,
527
- "supportsCacheControlOnTools": false,
528
- "supportsLongCacheRetention": false
480
+ "supportsReasoningEffort": true
529
481
  }
530
482
  },
531
483
  "accounts/fireworks/routers/kimi-k2p5-fast": {
@@ -544,20 +496,6 @@
544
496
  "supportsReasoningEffort": true
545
497
  }
546
498
  },
547
- "accounts/fireworks/routers/kimi-k2p5-turbo": {
548
- "name": "Kimi K2.5 Turbo (router)",
549
- "reasoning": true,
550
- "cost": {
551
- "input": 0.6,
552
- "output": 3,
553
- "cacheRead": 0.1,
554
- "cacheWrite": 0
555
- },
556
- "maxTokens": 256000,
557
- "compat": {
558
- "supportsReasoningEffort": true
559
- }
560
- },
561
499
  "accounts/fireworks/routers/kimi-k2p6": {
562
500
  "name": "Kimi K2.6 (router)",
563
501
  "reasoning": true,
@@ -576,8 +514,6 @@
576
514
  },
577
515
  "accounts/fireworks/routers/kimi-k2p6-fast": {
578
516
  "name": "Kimi K2.6 Fast",
579
- "api": "anthropic-messages",
580
- "baseUrl": "https://api.fireworks.ai/inference",
581
517
  "reasoning": true,
582
518
  "input": ["text", "image"],
583
519
  "cost": {
@@ -589,16 +525,11 @@
589
525
  "contextWindow": 262000,
590
526
  "maxTokens": 262000,
591
527
  "compat": {
592
- "sendSessionAffinityHeaders": true,
593
- "supportsEagerToolInputStreaming": false,
594
- "supportsCacheControlOnTools": false,
595
- "supportsLongCacheRetention": false
528
+ "supportsReasoningEffort": true
596
529
  }
597
530
  },
598
531
  "accounts/fireworks/routers/kimi-k2p6-turbo": {
599
532
  "name": "Kimi K2.6 Turbo",
600
- "api": "anthropic-messages",
601
- "baseUrl": "https://api.fireworks.ai/inference",
602
533
  "reasoning": true,
603
534
  "input": ["text", "image"],
604
535
  "cost": {
@@ -610,16 +541,11 @@
610
541
  "contextWindow": 262000,
611
542
  "maxTokens": 262000,
612
543
  "compat": {
613
- "sendSessionAffinityHeaders": true,
614
- "supportsEagerToolInputStreaming": false,
615
- "supportsCacheControlOnTools": false,
616
- "supportsLongCacheRetention": false
544
+ "supportsReasoningEffort": true
617
545
  }
618
546
  },
619
547
  "accounts/fireworks/routers/kimi-k2p7-code-fast": {
620
548
  "name": "Kimi K2.7 Code Fast",
621
- "api": "anthropic-messages",
622
- "baseUrl": "https://api.fireworks.ai/inference",
623
549
  "reasoning": true,
624
550
  "input": ["text", "image"],
625
551
  "cost": {
@@ -631,10 +557,7 @@
631
557
  "contextWindow": 262000,
632
558
  "maxTokens": 262000,
633
559
  "compat": {
634
- "sendSessionAffinityHeaders": true,
635
- "supportsEagerToolInputStreaming": false,
636
- "supportsCacheControlOnTools": false,
637
- "supportsLongCacheRetention": false
560
+ "supportsReasoningEffort": true
638
561
  }
639
562
  },
640
563
  "accounts/fireworks/routers/minimax-m2p7": {