syntaxmatrix 2.5.6__py3-none-any.whl → 2.6.2__py3-none-any.whl

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 (41) hide show
  1. syntaxmatrix/agentic/agents.py +1220 -169
  2. syntaxmatrix/agentic/agents_orchestrer.py +326 -0
  3. syntaxmatrix/agentic/code_tools_registry.py +27 -32
  4. syntaxmatrix/commentary.py +16 -16
  5. syntaxmatrix/core.py +185 -81
  6. syntaxmatrix/db.py +460 -4
  7. syntaxmatrix/{display.py → display_html.py} +2 -6
  8. syntaxmatrix/gpt_models_latest.py +1 -1
  9. syntaxmatrix/media/__init__.py +0 -0
  10. syntaxmatrix/media/media_pixabay.py +277 -0
  11. syntaxmatrix/models.py +1 -1
  12. syntaxmatrix/page_builder_defaults.py +183 -0
  13. syntaxmatrix/page_builder_generation.py +1122 -0
  14. syntaxmatrix/page_layout_contract.py +644 -0
  15. syntaxmatrix/page_patch_publish.py +1471 -0
  16. syntaxmatrix/preface.py +142 -21
  17. syntaxmatrix/profiles.py +28 -10
  18. syntaxmatrix/routes.py +1740 -453
  19. syntaxmatrix/selftest_page_templates.py +360 -0
  20. syntaxmatrix/settings/client_items.py +28 -0
  21. syntaxmatrix/settings/model_map.py +1022 -207
  22. syntaxmatrix/settings/prompts.py +328 -130
  23. syntaxmatrix/static/assets/hero-default.svg +22 -0
  24. syntaxmatrix/static/icons/bot-icon.png +0 -0
  25. syntaxmatrix/static/icons/favicon.png +0 -0
  26. syntaxmatrix/static/icons/logo.png +0 -0
  27. syntaxmatrix/static/icons/logo3.png +0 -0
  28. syntaxmatrix/templates/admin_branding.html +104 -0
  29. syntaxmatrix/templates/admin_features.html +63 -0
  30. syntaxmatrix/templates/admin_secretes.html +108 -0
  31. syntaxmatrix/templates/dashboard.html +296 -133
  32. syntaxmatrix/templates/dataset_resize.html +535 -0
  33. syntaxmatrix/templates/edit_page.html +2535 -0
  34. syntaxmatrix/utils.py +2431 -2383
  35. {syntaxmatrix-2.5.6.dist-info → syntaxmatrix-2.6.2.dist-info}/METADATA +6 -2
  36. {syntaxmatrix-2.5.6.dist-info → syntaxmatrix-2.6.2.dist-info}/RECORD +39 -24
  37. syntaxmatrix/generate_page.py +0 -644
  38. syntaxmatrix/static/icons/hero_bg.jpg +0 -0
  39. {syntaxmatrix-2.5.6.dist-info → syntaxmatrix-2.6.2.dist-info}/WHEEL +0 -0
  40. {syntaxmatrix-2.5.6.dist-info → syntaxmatrix-2.6.2.dist-info}/licenses/LICENSE.txt +0 -0
  41. {syntaxmatrix-2.5.6.dist-info → syntaxmatrix-2.6.2.dist-info}/top_level.txt +0 -0
syntaxmatrix/settings/model_map.py CHANGED
@@ -4,293 +4,1105 @@ import os
4
4
 
5
5
  PROVIDERS_MODELS = {
6
6
  #1
7
- "openai": [
8
- "gpt-5.1",
9
- "gpt-5.1-chat-latest",
10
- "gpt-5.1-codex-mini",
11
- "gpt-5.1-codex",
12
- "gpt-5",
13
- "gpt-5-nano",
14
- "gpt-5-mini",
15
- "gpt-4.1",
16
- "gpt-4.1-nano",
17
- "gpt-4.1-mini",
18
- "gpt-4o-mini",
19
- "gpt-4o",
7
+ "OpenAI": [
8
+ "gpt-5.2", # $1.75 $0.175 $14.00
9
+ "gpt-5.2-chat-latest", # $1.75 $0.175 $14.00
10
+ "gpt-5.2-pro", # $21.00 - $168.00
11
+ "gpt-5.1", # $1.25 $0.125 $10.00
12
+ "gpt-5.1-chat-latest", # $1.25 $0.125 $10.00
13
+ "gpt-5.1-codex-mini",
14
+ "gpt-5.1-codex-max", # $1.25 $0.125 $10.00
15
+ "gpt-5", # $1.25 $0.125 $10.00
16
+ "gpt-5-nano", # $0.05 $0.005 $0.40
17
+ "gpt-5-mini", # $0.25 $0.025 $2.00
18
+ "gpt-5-pro", # $15.00 - $120.00
19
+ "gpt-4.1", # $2.00 $0.50 $8.00
20
+ "gpt-4.1-nano", # $0.10 $0.025 $0.40
21
+ "gpt-4.1-mini", # $0.40 $0.10 $1.60
22
+ "gpt-4o", # $2.50 $1.25 $10.00
23
+ "gpt-4o-mini", # $0.15 $0.075 $0.60
24
+ # "gpt-4o-mini-search-preview", # $0.15 - $0.60
20
25
  ],
21
26
  #2
22
- "google": [
23
- "gemini-3-pro-preview",
27
+ "Google": [
28
+ "gemini-3-pro-preview",
29
+ "gemini-3-flash-preview",
30
+ "gemini-2.5-pro",
31
+ "gemini-2.5-flash",
24
32
  "gemini-2.5-flash-lite",
25
- "gemini-2.5-flash",
26
- "gemini-2.5-pro",
27
- "gemini-2.0-flash-lite",
28
33
  "gemini-2.0-flash",
34
+ "gemini-2.0-flash-lite",
35
+ # Image models
36
+ # "gemini-3-pro-image-preview",
37
+ # "gemini-2.5-flash-image",
38
+ # "imagen-4.0-generate-001",
39
+ # "imagen-4.0-ultra-generate-001",
40
+ # "imagen-4.0-fast-generate-001",
29
41
  ],
30
42
  #3
31
- "xai": [
32
- "grok-4",
33
- "grok-3-mini-fast",
34
- "grok-3-mini",
35
- "grok-3",
36
-
43
+ "xAI": [
44
+ "grok-4-1-fast-reasoning",
45
+ "grok-4-1-fast-non-reasoning",
46
+ "grok-4",
47
+ "grok-code-fast",
37
48
  ],
38
49
  #4
39
- "deepseek": [
40
- "deepseek-chat",
50
+ "Anthropic": [
51
+ "claude-opus-4-5",
52
+ "claude-sonnet-4-5",
53
+ "claude-4-5-haiku",
41
54
  ],
42
55
  #5
43
- "moonshot": [
44
- "kimi-k2-0905-preview",
56
+ "DeepSeek": [
57
+ "deepseek-reasoner",
58
+ "deepseek-chat",
45
59
  ],
46
60
  #6
47
- "alibaba": [
48
- "qwen-flash",
49
- "qwen-plus",
61
+ "Alibaba": [
62
+ "qwen3-max",
50
63
  "qwen3-coder-plus",
51
- "qwen-max",
64
+ "qwen3-coder-flash",
65
+ "qwen-plus",
66
+ "qwen-flash",
67
+
52
68
  ],
69
+
53
70
  #7
54
- "anthropic": [
55
- "claude-opus-4-5",
56
- "claude-opus-4-1",
57
- "claude-sonnet-4-5",
58
- "claude-sonnet-4-0",
59
- "claude-3-5-haiku-latest",
60
- "claude-3-haiku-20240307",
61
- ]
71
+ "MoonShot": [
72
+ "kimi-k2-0905-preview",
73
+ "kimi-k2-turbo-preview",
74
+ "kimi-k2-thinking",
75
+ "kimi-k2-thinking-turbo",
76
+ ],
62
77
  }
63
78
 
64
79
 
80
+ # #8
81
+ # "ZAI": [ # coding ==> https://api.z.ai/api/coding/paas/v4
82
+ # "glm-4.7",
83
+ # "glm-4.6", # general ==> https://api.z.ai/api/paas/v4
84
+ # "glm-4.6v",
85
+ # "glm-4.6v-flash",
86
+ # "glm-4.6v-flashx",
87
+ # "glm-4.5v",
88
+ # "glm-4-32b-0414-128k",
89
+ # "cogView-4-250304",
90
+ # "cogvideox-3",
91
+ # ]
92
+
93
+
65
94
  # Read-only model descriptions for LLM-profile builder
66
95
  # -----------------------------------------------------------------------------
67
96
  MODEL_DESCRIPTIONS = {
68
- #1. OpenAI
69
- "gpt-4o-mini":"Cost-efficient multimodal; $0.15/1M input, $0.60/1M output. Ideal for prototyping vision+text apps on a budget.",
97
+ #1.1 OpenAI
98
+ "gpt-4o":"""
99
+ Model: GPT 4o
100
+ Cost:
101
+ Input = $2.50 <= 1M tokens
102
+ Output = $10.00 <= 1M tokens
103
+
104
+ Data Type:
105
+ Input = (Text, Image)
106
+ Output = Text
107
+
108
+ Context Length:
109
+ Input = 128,000 tokens
110
+ Output = 16,384 tokens
111
+
112
+ Speed, Intelligence, and Training:
113
+ 3x Fast
114
+ 3x Clever
115
+ Cut-off: 01/10/2023
116
+
117
+ Agency:
118
+ - Admin
119
+ - Chat
120
+ - Classifier
121
+ - Summarizer
122
+ - ImageTexter
123
+ """,
124
+
125
+ #1.2 OpenAI
126
+ "gpt-4o-mini":"""
127
+ Model: GPT 4o Mini
128
+ Cost:
129
+ Input = $0.15 <= 1M tokens
130
+ Output = $0.60 <= 1M tokens
131
+
132
+ Data Type:
133
+ Input = (Text, Image)
134
+ Output = Text
70
135
 
71
- "gpt-4o":"Multimodal powerhouse; $5.00/1M input, $15.00/1M output. Best for high-fidelity chat, complex reasoning & image tasks.",
136
+ Context Length:
137
+ Input = 128,000 tokens
138
+ Output = 16,384 tokens
139
+
140
+ Speed, Intelligence, and Training:
141
+ 4x Fast
142
+ 2x Clever
143
+ Cut-off: 01/10/2023
144
+
145
+ Agency:
146
+ - Chat
147
+ - Classifier
148
+ - Summarizer
149
+ - ImageTexter
150
+ """,
72
151
 
73
- "gpt-4.1-nano":"Ultra-fast low-cost (1M-token); $0.10/1M in, $0.40/1M out. Perfect for high-throughput, low-latency tasks.",
152
+ #1.3 OpenAI
153
+ "gpt-4.1":"""
154
+ Model: GPT 4.1
155
+ Cost:
156
+ Input = $2.00 <= 1M tokens
157
+ Output = $8.00 <= 1M tokens
158
+
159
+ Data Type:
160
+ Input = (Text, Image)
161
+ Output = Text
162
+
163
+ Context Length:
164
+ Input = 1M tokens
165
+ Output = 32,768 tokens
166
+
167
+ Speed, Intelligence, and Training:
168
+ 3x Fast
169
+ 4x Clever
170
+ Knowledge: 01/06/2024
171
+
172
+ Agency:
173
+ - Coder
174
+ """,
175
+
176
+ #1.4 OpenAI
177
+ "gpt-4.1-nano":"""
178
+ Model: GPT 4.1 Nano
179
+ Cost:
180
+ Input = $0.10 <= 1M tokens
181
+ Output = $0.40 <= 1M tokens
182
+
183
+ Data Type:
184
+ Input = (Text, Image)
185
+ Output = Text
186
+
187
+ Context Length:
188
+ Input = 1M tokens
189
+ Output = 32,768 tokens
190
+
191
+ Speed, Intelligence, and Training:
192
+ 5x Fast
193
+ 2x Clever
194
+ Knowledge: 01/06/2024
195
+
196
+ Agency:
197
+ - Chat
198
+ - Classifier
199
+ - Summarizer
200
+ - ImageTexter
201
+ """,
74
202
 
75
- "gpt-4.1-mini":"Balanced speed/intel (1M-token context); $0.40/1M in, $1.60/1M out. Great for apps needing wide context at moderate cost.",
203
+ #1.5 OpenAI
204
+ "gpt-4.1-mini":"""
205
+ Model: GPT 4.1 Mini
206
+ Cost:
207
+ Input = $0.40 <= 1M tokens
208
+ Output = $1.60 <= 1M tokens
76
209
 
77
- "gpt-4.1":"Top general-purpose (1M-token context); $2.00/1M in, $8.00/1M out. Excels at large-doc comprehension, coding, reasoning.",
210
+ Data Type:
211
+ Input = (Text, Image)
212
+ Output = Text
78
213
 
79
- "gpt-5-chat-latest":"""gpt-5-main. """,
214
+ Context Length:
215
+ Input = 1M tokens
216
+ Output = 32,768 tokens
80
217
 
81
- "gpt-5-nano":"""gpt-5-thinking-nano. In/Out €0.043/€0.344 (cached in €0.004).
82
- Fastest/lowest cost; ideal for short prompts, tagging, and rewrite flows; tools supported.
83
- Best for:
84
- 1. High-volume classification/moderation
85
- 2. Copy clean-up and templated rewrites
86
- 3. Lightweight summarisation and routing
218
+ Speed, Intelligence, and Training:
219
+ 4x Fast
220
+ 3x Clever
221
+ Knowledge: 01/06/2024
87
222
 
88
- Use cases:
89
- a. Real-time content moderation and policy tagging.
90
- b. Bulk product description normalisation with style rules.
91
- c. News/article triage to decide which items warrant a deeper pass.
223
+ Agency:
224
+ - Admin
225
+ - Chat
226
+ - Classifier
227
+ - Summarizer
228
+ - ImageTexter
92
229
  """,
93
230
 
94
- "gpt-5-mini":"""gpt-5-thinking-mini. In/Out $0.25/$2 (cached in $0.025).
95
- Cheaper, faster variant with broad task coverage; still supports tools and long context.
96
- Best for:
97
- 1. Production chatbots at scale
98
- 2. Mid-complexity RAG/extraction pipelines
99
- 3. Batch summarisation with occasional tool calls
100
- Use cases:
101
- a. Customer support copilot that classifies intent, drafts replies, and calls ticketing APIs.
102
- b. Meeting-notes pipeline: diarised summary, actions, CRM updates.
103
- c. ETL enrichment: pull facts from documents into structured JSON.
231
+ #1.6 OpenAI
232
+ "gpt-5":"""
233
+ Model: GPT 5
234
+ Cost:
235
+ Input = $1.25 <= 1M tokens
236
+ Output = $10.00 <= 1M tokens
237
+
238
+ Data Type:
239
+ Input = (Text, Image)
240
+ Output = Text
241
+
242
+ Context Length:
243
+ Input = 400,000 tokens
244
+ Output = 128,000 tokens
245
+
246
+ Misc:
247
+ 3x Fast
248
+ 4x Clever
249
+ Thinking: Yes
250
+ Knowledge: 30/09/2024
251
+
252
+ Agency:
253
+ - Coder
254
+ """,
255
+
256
+ #1.7 OpenAI
257
+ "gpt-5-nano":"""
258
+ Model: GPT 5 Nano
259
+ Cost:
260
+ Input = $0.05 <= 1M tokens
261
+ Output = $0.40 <= 1M tokens
262
+
263
+ Data Type:
264
+ Input = (Text, Image)
265
+ Output = Text
266
+
267
+ Context Length:
268
+ Input = 400,000 tokens
269
+ Output = 128,000 tokens
270
+
271
+ Misc:
272
+ 5x Fast
273
+ 2x Clever
274
+ Thinking: Yes
275
+ Knowledge: 31/05/2024
276
+
277
+ Agency:
278
+ - Chat
279
+ - Classifier
280
+ - Summarizer
281
+ - ImageTexter
104
282
  """,
105
283
 
106
- "gpt-5":"""gpt-5-thinking. In/Out $1.25/$10.00 (cached in $0.125).
107
- Advanced reasoning and tool use; strong code generation/repair; robust long-context handling (400k).
108
- Best for:
109
- 1. Complex agentic workflows and planning
110
- 2. Long-context RAG and analytics
111
- 3. High-stakes coding assistance (multi-file changes & tests)
112
- Use cases:
113
- a. An autonomous “data room” analyst reading hundreds of PDFs and producing audit-ready briefs.
114
- b. A coding copilot that opens tickets, edits PRs, and runs tests via tools.</li>
115
- c. An enterprise chat assistant that reasons over policies and produces compliant outputs.
116
- """,
117
-
118
- # "gpt-o3":"High-accuracy reasoning (200K-token); $2.00/1M in, $8.00/1M out. Best for math, code gen, structured data outputs.",
119
- # "gpt-o4-mini":"Fast lean reasoning (200K-token); $1.10/1M in, $4.40/1M out. Ideal for vision+code when o3 is overkill.",
120
- # "gpt-o4-mini-high":"Enhanced mini-engine; $2.50/1M in (est.), $10.00/1M out (est.). Suited for interactive assistants with visual reasoning.",
121
-
122
- # Google
123
- "gemma-3n-e4b-it":"""Gemma is free.
124
- Best for: Use case:
125
- - Low latency | - Visual and text processing
126
- - Multilingual | - Text translation
127
- - Summarization | - Summarizing text research content
128
- """,
129
-
130
- #2 Google
131
- "gemma-3n-e4b-it": """
132
- Open source for local hosting
133
- """,
284
+ #1.8 OpenAI
285
+ "gpt-5-mini":"""
286
+ Model: GPT 5 Mini
287
+ Cost:
288
+ Input = $0.25 <= 1M tokens
289
+ Output = $2.00 <= 1M tokens
290
+
291
+ Data Type:
292
+ Input = (Text, Image)
293
+ Output = Text
134
294
 
135
- "gemini-2.0-flash-lite":"""$0.075 In, $0.30 Out. CoD: Aug 2024"
136
- Best for: Use case:
137
- - Long Context | - rocess 10,000 lines of code
138
- - Realtime streaming | - Call tools natively
139
- - Native tool use | - Stream images and video in realtime
295
+ Context Length:
296
+ Input = 400,000 tokens
297
+ Output = 128,000 tokens
298
+
299
+ Misc:
300
+ 4x Fast
301
+ 3x Clever
302
+ Thinking: Yes
303
+ Knowledge: 31/05/2024
304
+
305
+ Agency:
306
+ - Admin
307
+ - Chat
308
+ - Classifier
309
+ - Summarizer
310
+ - ImageTexter
311
+ """,
312
+
313
+ #1.7 OpenAI
314
+
315
+ #1.8 OpenAI
316
+
317
+ #1.8 OpenAI
318
+
319
+ #1.9 OpenAI
320
+ "gpt-5-pro":"""
321
+ Model: GPT 5 Pro
322
+ Cost:
323
+ Input = $15.00 <= 1M tokens
324
+ Output = $120.00 <= 1M tokens
325
+
326
+ Data Type:
327
+ Input = (Text, Image)
328
+ Output = Text
329
+
330
+ Context Length:
331
+ Input = 400,000 tokens
332
+ Output = 272,000 tokens
333
+
334
+ Misc:
335
+ 1x Fast
336
+ 5x Clever
337
+ Thinking: Yes
338
+ Knowledge: 30/09/2024
339
+
340
+ Agency:
341
+ - Coder
140
342
  """,
141
343
 
142
- "gemini-2.0-flash": """$0.10 In, $0.40 Out. CoD: Aug 2024
143
- Best for: Use case:
144
- - Multimodal understanding | - Process 10,000 lines of code
145
- - Realtime streaming | - Call tools natively, like Search
146
- - Native tool use | - Stream images & vids in R time
344
+ #1.10 OpenAI
345
+ "gpt-5.1":"""
346
+ Model: GPT 5.1
347
+ Cost:
348
+ Input = $1.25 <= 1M tokens
349
+ Output = $10.00 <= 1M tokens
350
+
351
+ Data Type:
352
+ Input = (Text, Image)
353
+ Output = Text
354
+
355
+ Context Length:
356
+ Input = 400,000 tokens
357
+ Output = 128,000 tokens
358
+
359
+ Misc:
360
+ 3x Fast
361
+ 4x Clever
362
+ Thinking: Yes
363
+ Knowledge: 30/09/2024
364
+
365
+ Agency:
366
+ - Admin
367
+ - Coder
147
368
  """,
148
369
 
149
- "gemini-2.5-flash-lite": "($0.10 In, $0.40 Out)/1M (est.) CoD: Jan 2025."
150
- " Best for: Use case:"
151
- " - Large scale processing - Data transformation"
152
- " - Low latency, high volume - Translation"
153
- " tasks with thinking - Summarizationt",
370
+ #1.10 OpenAI
371
+ "gpt-5.1-chat-latest":"""
372
+ Model: GPT 5.1 Chat
373
+ Cost:
374
+ Input = $1.25 <= 1M tokens
375
+ Output = $10.00 <= 1M tokens
376
+
377
+ Data Type:
378
+ Input = (Text, Image)
379
+ Output = Text
380
+
381
+ Context Length:
382
+ Input = 128,000 tokens
383
+ Output = 16,384 tokens
154
384
 
155
- "gemini-2.5-flash": """$0.30. $2.50 Out CoD: Jan 2024.
156
- Best for: Use case:
157
- - Large scale processing - Reason over complex problems
158
- - Low latency, high volume tasks - Show thinking process
159
- - Agentic use cases - Call tools natively
385
+ Misc:
386
+ 3x Fast
387
+ 3x Clever
388
+ Thinking: Yes
389
+ Knowledge: 30/09/2024
390
+
391
+ Agency:
392
+ - Admin
393
+ - Chat
160
394
  """,
161
395
 
162
- "gemini-2.5-pro": """$3.00 In /1M (est.). Advanced analytics, detailed reports & multi-step reasoning.
163
- Best for:
164
- - Coding
165
- - Reasoning
166
- - Multimodal understanding
396
+ #1.10 OpenAI
397
+ "gpt-5.1-codex-mini":"""
398
+ Model: GPT 5.1 Codex Mini
399
+ Cost:
400
+ Input = $0.25 <= 1M tokens
401
+ Output = $2.00 <= 1M tokens
402
+
403
+ Data Type:
404
+ Input = (Text, Image)
405
+ Output = Text
406
+
407
+ Context Length:
408
+ Input = 400,000 tokens
409
+ Output = 128,000 tokens
167
410
 
168
- Use case:
169
- - Reason over complex problems
170
- - Tackle difficult code, math and STEM problems
171
- - Use the long context for analyzing large datasets, codebases or documents
411
+ Misc:
412
+ 3x Fast
413
+ 4x Clever
414
+ Thinking: Yes
415
+ Knowledge: 30/09/2024
416
+
417
+ Agency:
418
+ - Coder
172
419
  """,
173
420
 
174
- #3 XAI
175
- "grok-3-mini-fast": "$0.20/1M (est.). "
176
- "Ultra-low latency chat, real-time monitoring & streaming apps.",
421
+ #1.11 OpenAI
422
+ "gpt-5.1-codex-max":"""
423
+ Model: GPT 5.1 Codex Max
424
+ Cost:
425
+ Input = $1.25 <= 1M tokens
426
+ Output = $10.00 <= 1M tokens
427
+
428
+ Data Type:
429
+ Input = (Text, Image)
430
+ Output = Text
431
+
432
+ Context Length:
433
+ Input = 400,000 tokens
434
+ Output = 128,000 tokens
435
+
436
+ Misc:
437
+ 4x Fast
438
+ 4x Clever
439
+ Thinking: Yes
440
+ Knowledge: 30/09/2024
441
+
442
+ Agency:
443
+ - Coder
444
+ """,
177
445
 
178
- "grok-3-mini": "$0.40/1M (est.). Budget-friendly chat & assistant tasks with good accuracy.",
446
+ #1.12 OpenAI
447
+ "gpt-5.2":"""
448
+ Model: GPT 5.2
449
+ Cost:
450
+ Input = $1.75 <= 1M tokens
451
+ Output = $14.00 <= 1M tokens
179
452
 
180
- "grok-3": "$1.00/1M (est.). General-purpose chat & content gen with balanced speed/quality.",
453
+ Data Type:
454
+ Input = (Text, Image)
455
+ Output = Text
181
456
 
182
- #4 DeepSeek
183
- "deepseek-chat": "DeepSeek Chat; $1.20/1M (est.). Optimized for private-data Q&A, enterprise search & document ingestion.",
457
+ Context Length:
458
+ Input = 400,000 tokens
459
+ Output = 128,000 tokens
184
460
 
185
- #5 MoonShot
186
- "kimi-k2-0905-preview": """Mixture-of-Experts (MoE). Context length of 256k.
187
- Enhanced Agentic Coding abilities, improved frontend code aesthetics and practicality, and better context understanding.
461
+ Misc:
462
+ 3x Fast
463
+ 5x Clever
464
+ Thinking: Yes
465
+ Knowledge: 31/08/2025
466
+
467
+ Agency:
468
+ - Coder
469
+ """,
470
+
471
+ #1.13 OpenAI
472
+ "gpt-5.2-pro":"""
473
+ Model: GPT 5.2 Pro
474
+ Cost:
475
+ Input = $21.00 <= 1M tokens
476
+ Output = $168.00 <= 1M tokens
477
+
478
+ Data Type:
479
+ Input = (Text, Image)
480
+ Output = Text
481
+
482
+ Context Length:
483
+ Input = 400,000 tokens
484
+ Output = 128,000 tokens
485
+
486
+ Misc:
487
+ 1x Fast
488
+ 5x Clever
489
+ Thinking: Yes
490
+ Knowledge: 31/08/2025
491
+
492
+ Agency:
493
+ - Coder
494
+ """,
495
+
496
+ #2.1 Google
497
+ "gemini-3-pro-preview": """
498
+ Model: Gemini 3 Pro
499
+ Cost:
500
+ Input = $2.00 <= 200k tokens / $4.00 > 200k tokens
501
+ Output = $12.00 <= 200k tokens / $18.00 > 200k tokens
502
+
503
+ Data Type:
504
+ Input = (Text, Image, Video, Audio, and PDF)
505
+ Output = Text
506
+
507
+ Context Length:
508
+ Input = 1M tokens
509
+ Output = 65.5k tokens
510
+
511
+ Agency:
512
+ - Coder
513
+ """,
188
514
 
189
- Pricing (per 1M tokens):
190
- Input: $0.15
191
- Cache: $0.60
192
- Output: $2.50
515
+ #2.2 Google
516
+ "gemini-3-flash-preview": """
517
+ Model: Gemini 3 Flash
518
+ Cost:
519
+ Input = $0.50 <= 1M tokens
520
+ Output = $3.00 <= 1M tokens
521
+
522
+ Data Type:
523
+ Input = (Text, Image, Video, Audio, PDF)
524
+ Output = Text
525
+
526
+ Context Length:
527
+ Input = 1M tokens
528
+ Output = 65.5k tokens
529
+
530
+ Agencies recommended:
531
+ - Coder
532
+ - Admin
533
+ - Chat
534
+ - Classifier
535
+ - Summarizer
536
+ - ImageTexter
537
+ """,
538
+
539
+ #2.3 Google
540
+ "gemini-2.5-pro": """
541
+ Model: Gemini 2.5 Pro
542
+ Cost:
543
+ Input = $1.25 <= 200k tokens / $10.00 > 200k tokens
544
+ Output = $2.50 > 200k tokens / $15.00 > 200k tokens
545
+
546
+ Data Type:
547
+ Input = (Text, Image, Video, Audio, PDF)
548
+ Output = Text
549
+
550
+ Context Length:
551
+ Input = 1M tokens
552
+ Output = 65.5k tokens
553
+
554
+ Agencies recommended:
555
+ - Coder
556
+ """,
557
+
558
+ #2.4 Google
559
+ "gemini-2.5-flash": """
560
+ Model: Gemini 2.5 Flash
561
+ Cost:
562
+ Input = $0.30 <= 1M tokens
563
+ Output = $2.50 <= 1M tokens
564
+
565
+ Data Type:
566
+ Input = (Text, Image, Video, Audio)
567
+ Output = Text
568
+
569
+ Context Length:
570
+ Input = 1M tokens
571
+ Output = 65.5k tokens
572
+
573
+ Agencies recommended:
574
+ - Admin
575
+ - Chat
576
+ - Classifier
577
+ - Summarizer
578
+ - ImageTexter
579
+ """,
580
+
581
+ #2.5 Google
582
+ "gemini-2.5-flash-lite": """
583
+ Model: Gemini 2.5 Flash Lite
584
+ Cost:
585
+ Input = $0.10 <= 1M tokens
586
+ Output = $0.40 <= 1M tokens
587
+
588
+ Data Type:
589
+ Input = (Text, image, video, audio, PDF)
590
+ Output = Text
591
+
592
+ Context Length:
593
+ Input = 1M tokens
594
+ Output = 65.5k tokens
595
+
596
+ Agencies recommended:
597
+ - Chat
598
+ - ImageTexter
599
+ - Classifier
600
+ - Summarizer
601
+ - ImageTexter
602
+ """,
603
+
604
+ #2.6 Google
605
+ "gemini-2.0-flash": """
606
+ Model: Gemini 2.0 Flash
607
+ Cost:
608
+ Input = $0.10 <= 1M tokens
609
+ Output = $0.40 <= 1M tokens
610
+
611
+ Data Type:
612
+ Input = (Text, Image, Video, Audio)
613
+ Output = Text
614
+
615
+ Context Length:
616
+ Input = 1M tokens
617
+ Output = 8k tokens
618
+
619
+ Agencies recommended:
620
+ - ImageTexter
621
+ - Summarizer
622
+ - Classifier
623
+ """,
624
+
625
+ #2.7 Google
626
+ "gemini-2.0-flash-lite": """
627
+ Model: Gemini 2.0 Flash Lite
628
+
629
+ Cost:
630
+ Input = $0.075 <= 1M tokens
631
+ Output = $0.30 <= 1M tokens
632
+
633
+ Data Type:
634
+ Input = (Text, Image, Video, Audio)
635
+ Output = Text
636
+
637
+ Context Length:
638
+ Input = 1M tokens
639
+ Output = 8k tokens
640
+
641
+ Agencies recommended:
642
+ - ImageTexter
643
+ - Summarizer
644
+ - Classifier
193
645
  """,
194
646
 
195
- #6 Alibaba
196
- #i
197
- "qwen-flash": """ Qwen-Flash is a lightweight, high-speed large language model from Alibaba Cloud, optimized for efficiency and cost-effectiveness.
647
+ #3.1 XAI
648
+ "grok-4-1-fast-reasoning": """
649
+ Model: Grok 4.1 Fast Thinking
650
+ Cost:
651
+ Input = $0.20 <= 1M tokens
652
+ Output = $0.50 <= 1M tokens
653
+
654
+ Data Type:
655
+ Input = (Text, Image)
656
+ Output = Text
198
657
 
199
- Pricing (per 1M tokens):
200
- Input: $0.05
201
- Output: $0.40
658
+ Context Length:
659
+ Input = 2M tokens
660
+ Output = ?
202
661
 
203
- Best for:
204
- > Simple, high-speed tasks requiring low latency.
205
- > Cost-sensitive applications where budget is a priority.
206
- > Scenarios demanding large context windows (supports up to 1M tokens).
662
+ Agencies recommended:
663
+ - Coder
664
+ - ImageTexter
665
+ """,
666
+
667
+ #3.2 XAI
668
+ "grok-4-1-fast-non-reasoning": """
669
+ Model: Grok 4.1 Fast
670
+ Cost:
671
+ Input = $0.20 <= 1M tokens
672
+ Output = $0.50 <= 1M tokens
673
+
674
+ Data Type:
675
+ Input = (Text, Image)
676
+ Output = Text
207
677
 
208
- Use cases:
209
- > Real-time chat and dialogue systems needing quick responses.
210
- > Large-scale text processing (e.g., summarization, polishing).
211
- > Prototyping and development where rapid iteration is key.
678
+ Context Length:
679
+ Input = 2M tokens
680
+ Output = ?
212
681
 
213
- Note: Lacks advanced reasoning features like "deep thinking" mode found in higher-tier Qwen models (e.g., Qwen-Plus).
682
+ Agencies recommended:
683
+ - Admin
684
+ - Chat
685
+ - Summarizer
686
+ - Classifier
687
+ - ImageTexter
214
688
  """,
215
689
 
216
- #ii
217
- "qwen-plus": """LLM offering a balance of performance, speed, and cost. It features a 131,072 token context window and supports both thinking and non-thinking modes for enhanced reasoning.
690
+ #3.3 XAI
691
+ "grok-code-fast-1": """
692
+ Model: Grok Code Fast
693
+ Cost:
694
+ Input = $0.20 <= 1M tokens
695
+ Output = $1.50 <= 1M tokens
218
696
 
219
- Pricing (per 1M tokens):
220
- Input: $0.40
221
- Output: $1.20
697
+ Data Type:
698
+ Input = Text
699
+ Output = Text
222
700
 
223
- Best for:
224
- > Moderately complex reasoning tasks due to its enhanced reasoning capabilities and thinking mode support 16.
225
- > Multilingual applications, with support for over 100 languages, including strong Chinese and English performance 12.
226
- > Cost-sensitive deployments requiring a balance of capability and affordability 13.
701
+ Context Length:
702
+ Input = 256k tokens
703
+ Output = ?
227
704
 
228
- Use cases:
229
- > Customer service automation (e.g., chatbots, virtual assistants) 26.
230
- > Content generation and summarization (e.g., marketing copy, document summarization) 236.
231
- > Code generation and tool-assisted tasks due to its agent capabilities and tool-calling support
705
+ Agencies recommended:
706
+ - Code
707
+ - Chat
708
+ - Summarizer
709
+ - Classifier
232
710
  """,
233
711
 
234
- #iii
235
- "qwen3-Coder-Plus": """A commercial, high-performance coding model optimized for agentic tasks like tool use, browser interaction, and long-context code generation.
236
-
237
- Pricing (per 1M tokens):
238
- Input $1 (0-32K tokens), $1.8 (32K-128K), $3 (128K-256K), $6 (256K-1M).
239
- Output $5 (0-32K), $9 (32K-128K), $15 (128K-256K), $60 (256K-1M)
712
+ #3.4 XAI
713
+ "grok-4": """
714
+ Model: Grok 4
715
+ Cost:
716
+ Input = $3.00 <= 1M tokens
717
+ Output = $15.00 <= 1M tokens
240
718
 
241
- Best for:
242
- > Repository-scale coding (handles large codebases with long context).
243
- > Agentic workflows (tool calling, multi-step environment interactions).
244
- > Real-world software engineering (debugging, refactoring, SWE-bench tasks).
719
+ Data Type:
720
+ Input = Text
721
+ Output = Text
245
722
 
246
- Use cases:
247
- > Automating complex coding tasks (e.g., full-stack app generation, data storytelling).
248
- > Debugging and refactoring (identifying bugs, improving code quality).
249
- > Multi-turn coding with feedback (iterative problem-solving with execution).
250
- > Consider this model if you need long-context, agentic coding capabilities comparable to Claude Sonnet 17. Avoid if budget constraints outweigh performance needs.
723
+ Context Length:
724
+ Input = 256k tokens
725
+ Output = ?
726
+
727
+ Agencies recommended:
728
+ - Code
251
729
  """,
252
730
 
253
- #iv
254
- "qwen-max": """Alibaba Cloud's flagship large-scale Mixture-of-Experts (MoE) model, pretrained on 20+ trillion tokens and refined with SFT/RLHF. Competes with top models like GPT-4o and Claude-3.5-Sonnet in benchmarks.
731
+ #4.1 Anthropic
732
+ "claude-opus-4-5": """
733
+ Model: Claude Opus 4.5
734
+ Cost:
735
+ Input = $5.00 <= 1M tokens
736
+ Output = $25.00 <= 1M tokens
737
+
738
+ Data Type:
739
+ Input = Text, Image
740
+ Output = Text
741
+
742
+ Context Length:
743
+ Input = 200k tokens
744
+ Output = 64k
745
+
746
+ Speed, Intelligence, and Training:
747
+ 3x Fast
748
+ 5x Clever
749
+ Thinking: Yes
750
+ Knowledge: May 2025
751
+
752
+ Agencies recommended:
753
+ - Coder
754
+ """,
255
755
 
256
- Pricing (per 1M tokens):
257
- Input: $1.60 - Output: $6.40
756
+ #4.2 Anthropic
757
+ "claude-sonnet-4-5": """
758
+ Model: Claude Sonnet 4.5
759
+ Cost:
760
+ Input = $3.00 <= 1M tokens
761
+ Output = $15.00 <= 1M tokens
258
762
 
259
- Best for:
260
- > Complex, multi-step reasoning tasks 113
261
- > Multilingual applications (supports 100+ languages)
262
- > Coding and tool-calling precision 111
763
+ Data Type:
764
+ Input = Text, Image
765
+ Output = Text
263
766
 
264
- Use cases:
265
- > Advanced coding assistance and debugging 310
266
- > High-quality content creation (e.g., documents, scripts)
267
- > Large-context analysis (32K token window) for documents or data
767
+ Context Length:
768
+ Input = 200k tokens
769
+ Output = 64k
268
770
 
771
+ Speed, Intelligence, and Training:
772
+ 4x Fast
773
+ 4x Clever
774
+ Thinking: Yes
775
+ Knowledge: Jan 2025
776
+
777
+ Agencies recommended:
778
+ - Code
269
779
  """,
270
780
 
271
- "claude-opus-4-1":""" $15 / MTok $18.75 / MTok $30 / MTok $1.50 / MTok $75 / MTok
781
+ #4.3 Anthropic
782
+ "claude-haiku-4-5":"""
783
+ Model: Claude Haiku 4.5
784
+ Cost:
785
+ Input = $1.00 <= 1M tokens
786
+ Output = $5.00 <= 1M tokens
787
+
788
+ Data Type:
789
+ Input = Text, Image
790
+ Output = Text
791
+
792
+ Context Length:
793
+ Input = 200k tokens
794
+ Output = 64k
795
+
796
+ Speed, Intelligence, and Training:
797
+ 5x Fast
798
+ 3x Clever
799
+ Thinking: Yes
800
+ Knowledge: Feb 2025
801
+
802
+ Agencies recommended:
803
+ - Admin
804
+ - Chat
805
+ - Code
272
806
  """,
807
+
808
+ #5.1 DeepSeek
809
+ "deepseek-chat":"""
810
+ Model: DeepSeek Chat
811
+
812
+ Cost:
813
+ Input = $0.28 <= 1M tokens
814
+ Output = $0.42 <= 1M tokens
815
+
816
+ Data Type:
817
+ Input = Text
818
+ Output = Text
819
+
820
+ Context Length:
821
+ Input = 128,000 tokens
822
+ Output = 4k-Default / 8k-Max tokens
273
823
 
274
- "claude-sonnet-4-0":""" $3 / MTok $3.75 / MTok $6 / MTok $0.30 / MTok $15 / MTok
824
+ Misc:
825
+ 4x Fast
826
+ 4x Clever
827
+ Thinking: No
828
+ Knowledge: 31/05/2024
829
+
830
+ Agency:
831
+ - Chat
832
+ - Classifier
833
+ - Summarizer
275
834
  """,
276
835
 
277
- "claude-haiku-3-5-latest":""" $0.80 / MTok $1 / MTok $1.6 / MTok $0.08 / MTok $4 / MTok
836
+ #5.2 DeepSeek
837
+ "deepseek-reasoner":"""
838
+ Model: DeepSeek Reasoner
839
+
840
+ Cost:
841
+ Input = $0.28 <= 1M tokens
842
+ Output = $0.42 <= 1M tokens
843
+
844
+ Data Type:
845
+ Input = Text
846
+ Output = Text
847
+
848
+ Context Length:
849
+ Input = 128,000 tokens
850
+ Output = 32,000-Default / 64,000-Max tokens
851
+
852
+ Misc:
853
+ 4x Fast
854
+ 4x Clever
855
+ Thinking: Yes
856
+ Knowledge: 31/05/2024
857
+
858
+ Agency:
859
+ - Coder
860
+ - Chat
861
+ - Classifier
862
+ - Summarizer
278
863
  """,
279
864
 
280
- "claude-3-haiku-20240307":""" $0.25 / MTok $0.30 / MTok $0.50 / MTok $0.03 / MTok $1.25 / MTok
865
+ #6.1 Alibaba
866
+ "qwen3-max":"""
867
+ Model: Qwen3 Max
868
+ Cost:
869
+ Input: $1.20 <= 32k tokens/ $2.40 32k-128k tokens/ $3 128k-256k tokens
870
+ Output: $6.00 <= 32k tokens/ $12 32k-128k tokens/ $15 128k-256k tokens
871
+
872
+ Data Type:
873
+ Input = Text
874
+ Output = Text
875
+
876
+ Context Length:
877
+ Input = 252k
878
+ Output = 64k
879
+
880
+ Misc:
881
+ 3x Fast
882
+ 5x Clever
883
+ Thinking: Available
884
+ Knowledge: ?
885
+
886
+ Agency:
887
+ - Code
281
888
  """,
282
889
 
283
- }
890
+ #6.2 Alibaba
891
+ "qwen3-coder-plus":"""
892
+ Model: Qwen3 Coder Plus
893
+ Cost:
894
+ Input: $1.00 <= 32k tokens/ $1.80 <= 128k tokens/ $6 <= 1M tokens
895
+ Output: $5.00 <= 32k tokens/ $9.00 <= 128k tokens/ $60.00 <= 1M tokens
896
+
897
+ Data Type:
898
+ Input = Text
899
+ Output = Text
284
900
 
901
+ Context Length:
902
+ Input = 1M tokens
903
+ Output = 64k tokens
904
+
905
+ Misc:
906
+ 2x Fast
907
+ 5x Clever
908
+ Thinking: Available
909
+ Knowledge: ?
910
+
911
+ Agency:
912
+ - Code
913
+ """,
914
+
915
+ #6.3 Alibaba
916
+ "qwen3-coder-flash":"""
917
+ Model: Qwen3 Coder Flash
918
+ Cost:
919
+ Input: $0.30 <= 32k tokens/ $0.50 <= 128k tokens/ $0.80 <= 256k tokens/ $1.6 <= 1M tokens
920
+ Output: $1.50 <= 32k tokens/ $2.50 <= 128k tokens/ $4.00 <= 256k tokens/ $9.6 <= 1M tokens
921
+
922
+ Data Type:
923
+ Input = Text
924
+ Output = Text
925
+
926
+ Context Length:
927
+ Input = 997k tokens
928
+ Output = 64k tokens
929
+
930
+ Misc:
931
+ 2x Fast
932
+ 5x Clever
933
+ Thinking: Available
934
+ Knowledge: ?
935
+
936
+ Agency:
937
+ - Code
938
+ """,
939
+
940
+ #6.4 Alibaba
941
+ "qwen-plush":"""
942
+ Model: Qwen Plus
943
+ Cost:
944
+ Input: $0.40 <= 256k tokens/ $1.2 <= 1M tokens
945
+ Output: $1.20 <= 256k tokens/ $3.6 <= 1M tokens
946
+
947
+ Data Type:
948
+ Input = Text
949
+ Output = Text
950
+
951
+ Context Length:
952
+ Input = 995k tokens
953
+ Output = 32k
954
+
955
+ Misc:
956
+ 3x Fast
957
+ 4x Clever
958
+ Thinking: Available
959
+ Knowledge: ?
960
+
961
+ Agency:
962
+ - Code
963
+ - Chat
964
+ - Classifier
965
+ - Summarizer
966
+ """,
967
+
968
+ #6.5 Alibaba
969
+ "qwen-flash":"""
970
+ Model: Qwen Flash
971
+ Cost:
972
+ Input: $0.05 <= 256k tokens/ $0.25 <= 1M tokens
973
+ Output: $0.40 <= 256k tokens/ $2.00 <= 1M tokens
974
+
975
+ Data Type:
976
+ Input = Text
977
+ Output = Text
978
+
979
+ Context Length:
980
+ Input = 995k tokens
981
+ Output = 32k tokens
982
+
983
+ Misc:
984
+ 5x Fast
985
+ 3x Clever
986
+ Thinking: Available
987
+ Knowledge: ?
988
+
989
+ Agency:
990
+ - Chat
991
+ - Classifier
992
+ - Summarizer
993
+ """,
994
+
995
+ #7.1 Moonshot
996
+ "kimi-k2-0905-preview":"""
997
+ Model: Kimi K2
998
+ Cost:
999
+ Input = $0.15 <= 1M tokens
1000
+ Output = $2.50 <= 1M tokens
1001
+
1002
+ Data Type:
1003
+ Input = Text
1004
+ Output = Text
1005
+
1006
+ Context Length:
1007
+ Window: 256k tokens
1008
+
1009
+ Misc:
1010
+ 4x Fast
1011
+ 3x Clever
1012
+
1013
+ Agency:
1014
+ - Chat
1015
+ - Classifier
1016
+ - Summarizer
1017
+ """,
1018
+
1019
+
1020
+ #7.2 Moonshot
1021
+ "kimi-k2-turbo-preview":"""
1022
+ Model: Kimi K2 Turbo
1023
+ Cost:
1024
+ Input = $0.15 <= 1M tokens
1025
+ Output = $8.00 <= 1M tokens
1026
+
1027
+ Data Type:
1028
+ Input = Text
1029
+ Output = Text
1030
+
1031
+ Context Length:
1032
+ Window: 256k tokens
1033
+
1034
+ Misc:
1035
+ 5x Fast
1036
+ 3x Clever
1037
+
1038
+ Agency:
1039
+ - Chat
1040
+ - Classifier
1041
+ - Summarizer
1042
+ """,
1043
+
1044
+ #7.3 Moonshot
1045
+ "kimi-k2-thinking":"""
1046
+ Model: Kimi K2 Thinking
1047
+ Cost:
1048
+ Input = $0.15 <= 1M tokens
1049
+ Output = $2.50 <= 1M tokens
1050
+
1051
+ Data Type:
1052
+ Input = (Text, Image)
1053
+ Output = Text
1054
+
1055
+ Context Length:
1056
+ Window: 256k tokens
1057
+
1058
+ Misc:
1059
+ 3x Fast
1060
+ 4x Clever
1061
+
1062
+ Agency:
1063
+ - Coder
1064
+ - Chat
1065
+ - Classifier
1066
+ - Summarizer
1067
+ """,
1068
+
1069
+ #7.4 Moonshot
1070
+ "kimi-k2-thinking-turbo":"""
1071
+ Model: Kimi K2 Thinking Turbo
1072
+ Cost:
1073
+ Input = $0.15 <= 1M tokens
1074
+ Output = $8.00 <= 1M tokens
1075
+
1076
+ Data Type:
1077
+ Input = Text
1078
+ Output = Text
1079
+
1080
+ Context Length:
1081
+ Window: 256k tokens
1082
+
1083
+ Misc:
1084
+ 4x Fast
1085
+ 5x Clever
1086
+
1087
+ Agency:
1088
+ - Coder
1089
+ - Chat
1090
+ - Classifier
1091
+ - Summarizer
1092
+ """,
1093
+ }
285
1094
 
1095
+ # Agencies
286
1096
  # -----------------------------------------------------------------------------
287
1097
  PURPOSE_TAGS = [
288
1098
  "admin",
289
1099
  "chat",
290
- "coding",
291
- "vision2text",
292
- "classification",
293
- "summarization",
1100
+ "coder",
1101
+ "classifier",
1102
+ "summarizer",
1103
+ "imagetexter",
1104
+ "textimager",
1105
+ "imageeditor"
294
1106
  ]
295
1107
 
296
1108
  # -----------------------------------------------------------------------------
@@ -303,11 +1115,14 @@ EMBEDDING_MODELS = {
303
1115
 
304
1116
 
305
1117
  GPT_MODELS_LATEST = [
306
- "gpt-5.1",
307
- "gpt-5.1-chat-latest",
308
- "gpt-5.1-codex-mini",
309
- "gpt-5.1-codex",
1118
+ "gpt-5.2",
1119
+ "gpt-5.2-chat-latest", # $1.75 $0.175 $14.00
1120
+ "gpt-5.2-pro", # $21.00 - $168.00
1121
+ "gpt-5.1", # $1.25 $0.125 $10.00
1122
+ "gpt-5.1-chat-latest", # $1.25 $0.125 $10.00
1123
+ "gpt-5.1-codex-mini",
1124
+ "gpt-5.1-codex-max",
310
1125
  "gpt-5",
311
1126
  "gpt-5-nano",
312
- "gpt-5-mini",
1127
+ "gpt-5-mini",
313
1128
  ]