pi-free 2.2.4 → 2.2.6

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 (35) hide show
  1. package/CHANGELOG.md +24 -0
  2. package/README.md +1 -1
  3. package/banner.svg +1 -1
  4. package/config.ts +58 -29
  5. package/constants.ts +1 -4
  6. package/index.ts +30 -24
  7. package/lib/built-in-toggle.ts +185 -18
  8. package/lib/provider-cache.ts +22 -0
  9. package/package.json +74 -74
  10. package/provider-failover/benchmark-lookup.ts +4 -1
  11. package/provider-failover/benchmarks-chunk-0.ts +570 -570
  12. package/provider-failover/benchmarks-chunk-1.ts +676 -676
  13. package/provider-failover/benchmarks-chunk-2.ts +673 -673
  14. package/provider-failover/benchmarks-chunk-3.ts +680 -680
  15. package/provider-failover/benchmarks-chunk-4.ts +683 -683
  16. package/provider-failover/benchmarks-chunk-5.ts +816 -474
  17. package/provider-helper.ts +64 -0
  18. package/providers/bai/bai.ts +8 -2
  19. package/providers/crofai/crofai.ts +8 -2
  20. package/providers/deepinfra/deepinfra.ts +8 -2
  21. package/providers/dynamic-built-in/index.ts +36 -26
  22. package/providers/kilo/kilo.ts +41 -22
  23. package/providers/novita/novita.ts +8 -2
  24. package/providers/openmodel/openmodel.ts +8 -2
  25. package/providers/qoder/models.ts +63 -175
  26. package/providers/qoder/qoder.ts +49 -84
  27. package/providers/qoder/stream.ts +182 -274
  28. package/providers/qoder/transform.ts +5 -2
  29. package/providers/routeway/routeway.ts +8 -2
  30. package/providers/sambanova/sambanova.ts +12 -6
  31. package/providers/together/together.ts +8 -2
  32. package/providers/tokenrouter/tokenrouter.ts +8 -2
  33. package/providers/zenmux/zenmux.ts +8 -2
  34. package/providers/codestral/codestral.ts +0 -128
  35. package/providers/qoder/encoding.ts +0 -48
package/provider-failover/benchmarks-chunk-5.ts CHANGED
@@ -1,20 +1,20 @@
1
1
  // Auto-generated benchmark data chunk 5
2
- // Models: glm-4.5-air .. seed-oss-36b-instruct (68 entries)
3
- // Last updated: 2026-06-01
2
+ // Models: olmo-2-32b .. seed-oss-36b-instruct (86 entries)
3
+ // Last updated: 2026-07-01
4
4
  // DO NOT EDIT MANUALLY — generated by scripts/update-benchmarks.ts
5
5
 
6
6
  import type { HardcodedBenchmark } from "./hardcoded-benchmarks.ts";
7
7
 
8
8
  export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
9
- "glm-4.5-air": {
9
+ "olmo-2-32b": {
10
10
  // AA specific benchmarks
11
- codingIndex: 23.8,
12
- mathIndex: 80.7,
11
+ codingIndex: undefined,
12
+ mathIndex: 3.3,
13
13
 
14
14
  // Academic benchmarks
15
- mmluPro: 0.815,
16
- gpqa: 0.733,
17
- hle: 0.068,
15
+ mmluPro: 0.511,
16
+ gpqa: 0.328,
17
+ hle: 0.037,
18
18
 
19
19
  // Capabilities
20
20
  contextWindow: 8192,
@@ -22,18 +22,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
22
22
  supportsVision: false,
23
23
 
24
24
  // Metadata
25
- lastUpdated: "2026-06-01",
26
- originalModel: "GLM-4.5-Air",
25
+ lastUpdated: "2026-07-01",
26
+ originalModel: "OLMo 2 32B",
27
27
  },
28
- "glm-4.6v-reasoning": {
28
+ "granite-3.3-8b-non-reasoning": {
29
29
  // AA specific benchmarks
30
- codingIndex: 19.7,
31
- mathIndex: 85.3,
30
+ codingIndex: undefined,
31
+ mathIndex: 6.7,
32
32
 
33
33
  // Academic benchmarks
34
- mmluPro: 0.799,
35
- gpqa: 0.719,
36
- hle: 0.089,
34
+ mmluPro: 0.468,
35
+ gpqa: 0.338,
36
+ hle: 0.042,
37
37
 
38
38
  // Capabilities
39
39
  contextWindow: 8192,
@@ -41,18 +41,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
41
41
  supportsVision: false,
42
42
 
43
43
  // Metadata
44
- lastUpdated: "2026-06-01",
45
- originalModel: "GLM-4.6V (Reasoning)",
44
+ lastUpdated: "2026-07-01",
45
+ originalModel: "Granite 3.3 8B (Non-reasoning)",
46
46
  },
47
- "glm-4.5v-non-reasoning": {
47
+ "reka-flash-sep-24": {
48
48
  // AA specific benchmarks
49
- codingIndex: 10.8,
50
- mathIndex: 15.3,
49
+ codingIndex: undefined,
50
+ mathIndex: undefined,
51
51
 
52
52
  // Academic benchmarks
53
- mmluPro: 0.751,
54
- gpqa: 0.573,
55
- hle: 0.036,
53
+ mmluPro: undefined,
54
+ gpqa: undefined,
55
+ hle: undefined,
56
56
 
57
57
  // Capabilities
58
58
  contextWindow: 8192,
@@ -60,18 +60,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
60
60
  supportsVision: false,
61
61
 
62
62
  // Metadata
63
- lastUpdated: "2026-06-01",
64
- originalModel: "GLM-4.5V (Non-reasoning)",
63
+ lastUpdated: "2026-07-01",
64
+ originalModel: "Reka Flash (Sep '24)",
65
65
  },
66
- "command-r+-apr-24": {
66
+ "hermes-3---llama-3.1-70b": {
67
67
  // AA specific benchmarks
68
68
  codingIndex: undefined,
69
69
  mathIndex: undefined,
70
70
 
71
71
  // Academic benchmarks
72
- mmluPro: 0.432,
73
- gpqa: 0.323,
74
- hle: 0.045,
72
+ mmluPro: 0.571,
73
+ gpqa: 0.401,
74
+ hle: 0.041,
75
75
 
76
76
  // Capabilities
77
77
  contextWindow: 8192,
@@ -79,18 +79,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
79
79
  supportsVision: false,
80
80
 
81
81
  // Metadata
82
- lastUpdated: "2026-06-01",
83
- originalModel: "Command-R+ (Apr '24)",
82
+ lastUpdated: "2026-07-01",
83
+ originalModel: "Hermes 3 - Llama-3.1 70B",
84
84
  },
85
- "command-r-mar-24": {
85
+ "mimo-v2-pro": {
86
86
  // AA specific benchmarks
87
87
  codingIndex: undefined,
88
88
  mathIndex: undefined,
89
89
 
90
90
  // Academic benchmarks
91
- mmluPro: 0.338,
92
- gpqa: 0.284,
93
- hle: 0.048,
91
+ mmluPro: undefined,
92
+ gpqa: 0.87,
93
+ hle: 0.283,
94
94
 
95
95
  // Capabilities
96
96
  contextWindow: 8192,
@@ -98,18 +98,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
98
98
  supportsVision: false,
99
99
 
100
100
  // Metadata
101
- lastUpdated: "2026-06-01",
102
- originalModel: "Command-R (Mar '24)",
101
+ lastUpdated: "2026-07-01",
102
+ originalModel: "MiMo-V2-Pro",
103
103
  },
104
- "apriel-v1.5-15b-thinker": {
104
+ "mimo-v2-flash-reasoning": {
105
105
  // AA specific benchmarks
106
- codingIndex: 18.7,
107
- mathIndex: 87.5,
106
+ codingIndex: undefined,
107
+ mathIndex: 96.3,
108
108
 
109
109
  // Academic benchmarks
110
- mmluPro: 0.773,
111
- gpqa: 0.713,
112
- hle: 0.12,
110
+ mmluPro: 0.843,
111
+ gpqa: 0.846,
112
+ hle: 0.211,
113
113
 
114
114
  // Capabilities
115
115
  contextWindow: 8192,
@@ -117,18 +117,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
117
117
  supportsVision: false,
118
118
 
119
119
  // Metadata
120
- lastUpdated: "2026-06-01",
121
- originalModel: "Apriel-v1.5-15B-Thinker",
120
+ lastUpdated: "2026-07-01",
121
+ originalModel: "MiMo-V2-Flash (Reasoning)",
122
122
  },
123
- "jamba-1.5-large": {
123
+ "sarvam-m-reasoning": {
124
124
  // AA specific benchmarks
125
125
  codingIndex: undefined,
126
126
  mathIndex: undefined,
127
127
 
128
128
  // Academic benchmarks
129
- mmluPro: 0.572,
130
- gpqa: 0.427,
131
- hle: 0.04,
129
+ mmluPro: 0.696,
130
+ gpqa: 0.416,
131
+ hle: 0.033,
132
132
 
133
133
  // Capabilities
134
134
  contextWindow: 8192,
@@ -136,18 +136,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
136
136
  supportsVision: false,
137
137
 
138
138
  // Metadata
139
- lastUpdated: "2026-06-01",
140
- originalModel: "Jamba 1.5 Large",
139
+ lastUpdated: "2026-07-01",
140
+ originalModel: "Sarvam M (Reasoning)",
141
141
  },
142
- "jamba-1.5-mini": {
142
+ "glm-4.7-non-reasoning": {
143
143
  // AA specific benchmarks
144
144
  codingIndex: undefined,
145
- mathIndex: undefined,
145
+ mathIndex: 48,
146
146
 
147
147
  // Academic benchmarks
148
- mmluPro: 0.371,
149
- gpqa: 0.302,
150
- hle: 0.051,
148
+ mmluPro: 0.794,
149
+ gpqa: 0.664,
150
+ hle: 0.061,
151
151
 
152
152
  // Capabilities
153
153
  contextWindow: 8192,
@@ -155,18 +155,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
155
155
  supportsVision: false,
156
156
 
157
157
  // Metadata
158
- lastUpdated: "2026-06-01",
159
- originalModel: "Jamba 1.5 Mini",
158
+ lastUpdated: "2026-07-01",
159
+ originalModel: "GLM-4.7 (Non-reasoning)",
160
160
  },
161
- "jamba-1.6-mini": {
161
+ "glm-4.7-reasoning": {
162
162
  // AA specific benchmarks
163
163
  codingIndex: undefined,
164
- mathIndex: undefined,
164
+ mathIndex: 95,
165
165
 
166
166
  // Academic benchmarks
167
- mmluPro: 0.367,
168
- gpqa: 0.3,
169
- hle: 0.046,
167
+ mmluPro: 0.856,
168
+ gpqa: 0.859,
169
+ hle: 0.251,
170
170
 
171
171
  // Capabilities
172
172
  contextWindow: 8192,
@@ -174,18 +174,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
174
174
  supportsVision: false,
175
175
 
176
176
  // Metadata
177
- lastUpdated: "2026-06-01",
178
- originalModel: "Jamba 1.6 Mini",
177
+ lastUpdated: "2026-07-01",
178
+ originalModel: "GLM-4.7 (Reasoning)",
179
179
  },
180
- "jamba-1.6-large": {
180
+ "glm-4.6-reasoning": {
181
181
  // AA specific benchmarks
182
182
  codingIndex: undefined,
183
- mathIndex: undefined,
183
+ mathIndex: 86,
184
184
 
185
185
  // Academic benchmarks
186
- mmluPro: 0.565,
187
- gpqa: 0.387,
188
- hle: 0.04,
186
+ mmluPro: 0.829,
187
+ gpqa: 0.78,
188
+ hle: 0.133,
189
189
 
190
190
  // Capabilities
191
191
  contextWindow: 8192,
@@ -193,18 +193,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
193
193
  supportsVision: false,
194
194
 
195
195
  // Metadata
196
- lastUpdated: "2026-06-01",
197
- originalModel: "Jamba 1.6 Large",
196
+ lastUpdated: "2026-07-01",
197
+ originalModel: "GLM-4.6 (Reasoning)",
198
198
  },
199
- "arctic-instruct": {
199
+ "glm-4.6-non-reasoning": {
200
200
  // AA specific benchmarks
201
201
  codingIndex: undefined,
202
- mathIndex: undefined,
202
+ mathIndex: 44.3,
203
203
 
204
204
  // Academic benchmarks
205
- mmluPro: undefined,
206
- gpqa: undefined,
207
- hle: undefined,
205
+ mmluPro: 0.784,
206
+ gpqa: 0.632,
207
+ hle: 0.052,
208
208
 
209
209
  // Capabilities
210
210
  contextWindow: 8192,
@@ -212,18 +212,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
212
212
  supportsVision: false,
213
213
 
214
214
  // Metadata
215
- lastUpdated: "2026-06-01",
216
- originalModel: "Arctic Instruct",
215
+ lastUpdated: "2026-07-01",
216
+ originalModel: "GLM-4.6 (Non-reasoning)",
217
217
  },
218
- "qwen2.5-max": {
218
+ "glm-4.6v-non-reasoning": {
219
219
  // AA specific benchmarks
220
220
  codingIndex: undefined,
221
- mathIndex: undefined,
221
+ mathIndex: 26.3,
222
222
 
223
223
  // Academic benchmarks
224
- mmluPro: 0.762,
225
- gpqa: 0.587,
226
- hle: 0.045,
224
+ mmluPro: 0.752,
225
+ gpqa: 0.566,
226
+ hle: 0.037,
227
227
 
228
228
  // Capabilities
229
229
  contextWindow: 8192,
@@ -231,18 +231,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
231
231
  supportsVision: false,
232
232
 
233
233
  // Metadata
234
- lastUpdated: "2026-06-01",
235
- originalModel: "Qwen2.5 Max",
234
+ lastUpdated: "2026-07-01",
235
+ originalModel: "GLM-4.6V (Non-reasoning)",
236
236
  },
237
- "qwen2.5-instruct-72b": {
237
+ "glm-5-reasoning": {
238
238
  // AA specific benchmarks
239
- codingIndex: 11.9,
240
- mathIndex: 14,
239
+ codingIndex: undefined,
240
+ mathIndex: undefined,
241
241
 
242
242
  // Academic benchmarks
243
- mmluPro: 0.72,
244
- gpqa: 0.491,
245
- hle: 0.042,
243
+ mmluPro: undefined,
244
+ gpqa: 0.82,
245
+ hle: 0.272,
246
246
 
247
247
  // Capabilities
248
248
  contextWindow: 8192,
@@ -250,18 +250,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
250
250
  supportsVision: false,
251
251
 
252
252
  // Metadata
253
- lastUpdated: "2026-06-01",
254
- originalModel: "Qwen2.5 Instruct 72B",
253
+ lastUpdated: "2026-07-01",
254
+ originalModel: "GLM-5 (Reasoning)",
255
255
  },
256
- "qwen2.5-coder-instruct-32b": {
256
+ "glm-4.5v-reasoning": {
257
257
  // AA specific benchmarks
258
258
  codingIndex: undefined,
259
- mathIndex: undefined,
259
+ mathIndex: 73,
260
260
 
261
261
  // Academic benchmarks
262
- mmluPro: 0.635,
263
- gpqa: 0.417,
264
- hle: 0.038,
262
+ mmluPro: 0.788,
263
+ gpqa: 0.684,
264
+ hle: 0.059,
265
265
 
266
266
  // Capabilities
267
267
  contextWindow: 8192,
@@ -269,18 +269,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
269
269
  supportsVision: false,
270
270
 
271
271
  // Metadata
272
- lastUpdated: "2026-06-01",
273
- originalModel: "Qwen2.5 Coder Instruct 32B",
272
+ lastUpdated: "2026-07-01",
273
+ originalModel: "GLM-4.5V (Reasoning)",
274
274
  },
275
- "qwen2.5-turbo": {
275
+ "glm-4.7-flash-reasoning": {
276
276
  // AA specific benchmarks
277
277
  codingIndex: undefined,
278
278
  mathIndex: undefined,
279
279
 
280
280
  // Academic benchmarks
281
- mmluPro: 0.633,
282
- gpqa: 0.41,
283
- hle: 0.042,
281
+ mmluPro: undefined,
282
+ gpqa: 0.581,
283
+ hle: 0.071,
284
284
 
285
285
  // Capabilities
286
286
  contextWindow: 8192,
@@ -288,18 +288,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
288
288
  supportsVision: false,
289
289
 
290
290
  // Metadata
291
- lastUpdated: "2026-06-01",
292
- originalModel: "Qwen2.5 Turbo",
291
+ lastUpdated: "2026-07-01",
292
+ originalModel: "GLM-4.7-Flash (Reasoning)",
293
293
  },
294
- "qwen2-instruct-72b": {
294
+ "glm-4.5-reasoning": {
295
295
  // AA specific benchmarks
296
296
  codingIndex: undefined,
297
- mathIndex: undefined,
297
+ mathIndex: 73.7,
298
298
 
299
299
  // Academic benchmarks
300
- mmluPro: 0.622,
301
- gpqa: 0.371,
302
- hle: 0.037,
300
+ mmluPro: 0.835,
301
+ gpqa: 0.782,
302
+ hle: 0.122,
303
303
 
304
304
  // Capabilities
305
305
  contextWindow: 8192,
@@ -307,18 +307,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
307
307
  supportsVision: false,
308
308
 
309
309
  // Metadata
310
- lastUpdated: "2026-06-01",
311
- originalModel: "Qwen2 Instruct 72B",
310
+ lastUpdated: "2026-07-01",
311
+ originalModel: "GLM-4.5 (Reasoning)",
312
312
  },
313
- "qwq-32b": {
313
+ "glm-4.5v-non-reasoning": {
314
314
  // AA specific benchmarks
315
315
  codingIndex: undefined,
316
- mathIndex: 29,
316
+ mathIndex: 15.3,
317
317
 
318
318
  // Academic benchmarks
319
- mmluPro: 0.764,
320
- gpqa: 0.593,
321
- hle: 0.082,
319
+ mmluPro: 0.751,
320
+ gpqa: 0.573,
321
+ hle: 0.036,
322
322
 
323
323
  // Capabilities
324
324
  contextWindow: 8192,
@@ -326,18 +326,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
326
326
  supportsVision: false,
327
327
 
328
328
  // Metadata
329
- lastUpdated: "2026-06-01",
330
- originalModel: "QwQ 32B",
329
+ lastUpdated: "2026-07-01",
330
+ originalModel: "GLM-4.5V (Non-reasoning)",
331
331
  },
332
- "qwen3-vl-235b-a22b-instruct": {
332
+ "glm-5-non-reasoning": {
333
333
  // AA specific benchmarks
334
- codingIndex: 16.5,
335
- mathIndex: 70.7,
334
+ codingIndex: undefined,
335
+ mathIndex: undefined,
336
336
 
337
337
  // Academic benchmarks
338
- mmluPro: 0.823,
339
- gpqa: 0.712,
340
- hle: 0.063,
338
+ mmluPro: undefined,
339
+ gpqa: 0.666,
340
+ hle: 0.072,
341
341
 
342
342
  // Capabilities
343
343
  contextWindow: 8192,
@@ -345,18 +345,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
345
345
  supportsVision: false,
346
346
 
347
347
  // Metadata
348
- lastUpdated: "2026-06-01",
349
- originalModel: "Qwen3 VL 235B A22B Instruct",
348
+ lastUpdated: "2026-07-01",
349
+ originalModel: "GLM-5 (Non-reasoning)",
350
350
  },
351
- "qwen3-coder-30b-a3b-instruct": {
351
+ "glm-4.6v-reasoning": {
352
352
  // AA specific benchmarks
353
- codingIndex: 19.4,
354
- mathIndex: 29,
353
+ codingIndex: undefined,
354
+ mathIndex: 85.3,
355
355
 
356
356
  // Academic benchmarks
357
- mmluPro: 0.706,
358
- gpqa: 0.516,
359
- hle: 0.04,
357
+ mmluPro: 0.799,
358
+ gpqa: 0.719,
359
+ hle: 0.089,
360
360
 
361
361
  // Capabilities
362
362
  contextWindow: 8192,
@@ -364,18 +364,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
364
364
  supportsVision: false,
365
365
 
366
366
  // Metadata
367
- lastUpdated: "2026-06-01",
368
- originalModel: "Qwen3 Coder 30B A3B Instruct",
367
+ lastUpdated: "2026-07-01",
368
+ originalModel: "GLM-4.6V (Reasoning)",
369
369
  },
370
- "qwen3-4b-non-reasoning": {
370
+ "glm-4.5-air": {
371
371
  // AA specific benchmarks
372
372
  codingIndex: undefined,
373
- mathIndex: undefined,
373
+ mathIndex: 80.7,
374
374
 
375
375
  // Academic benchmarks
376
- mmluPro: 0.586,
377
- gpqa: 0.398,
378
- hle: 0.037,
376
+ mmluPro: 0.815,
377
+ gpqa: 0.733,
378
+ hle: 0.068,
379
379
 
380
380
  // Capabilities
381
381
  contextWindow: 8192,
@@ -383,18 +383,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
383
383
  supportsVision: false,
384
384
 
385
385
  // Metadata
386
- lastUpdated: "2026-06-01",
387
- originalModel: "Qwen3 4B (Non-reasoning)",
386
+ lastUpdated: "2026-07-01",
387
+ originalModel: "GLM-4.5-Air",
388
388
  },
389
- "qwen3-235b-a22b-2507-reasoning": {
389
+ "glm-4.7-flash-non-reasoning": {
390
390
  // AA specific benchmarks
391
- codingIndex: 23.2,
392
- mathIndex: 91,
391
+ codingIndex: undefined,
392
+ mathIndex: undefined,
393
393
 
394
394
  // Academic benchmarks
395
- mmluPro: 0.843,
396
- gpqa: 0.79,
397
- hle: 0.15,
395
+ mmluPro: undefined,
396
+ gpqa: 0.452,
397
+ hle: 0.049,
398
398
 
399
399
  // Capabilities
400
400
  contextWindow: 8192,
@@ -402,18 +402,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
402
402
  supportsVision: false,
403
403
 
404
404
  // Metadata
405
- lastUpdated: "2026-06-01",
406
- originalModel: "Qwen3 235B A22B 2507 (Reasoning)",
405
+ lastUpdated: "2026-07-01",
406
+ originalModel: "GLM-4.7-Flash (Non-reasoning)",
407
407
  },
408
- "qwen3-235b-a22b-2507-instruct": {
408
+ "command-r+-apr-24": {
409
409
  // AA specific benchmarks
410
- codingIndex: 22.1,
411
- mathIndex: 71.7,
410
+ codingIndex: undefined,
411
+ mathIndex: undefined,
412
412
 
413
413
  // Academic benchmarks
414
- mmluPro: 0.828,
415
- gpqa: 0.753,
416
- hle: 0.106,
414
+ mmluPro: 0.432,
415
+ gpqa: 0.323,
416
+ hle: 0.045,
417
417
 
418
418
  // Capabilities
419
419
  contextWindow: 8192,
@@ -421,18 +421,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
421
421
  supportsVision: false,
422
422
 
423
423
  // Metadata
424
- lastUpdated: "2026-06-01",
425
- originalModel: "Qwen3 235B A22B 2507 Instruct",
424
+ lastUpdated: "2026-07-01",
425
+ originalModel: "Command-R+ (Apr '24)",
426
426
  },
427
- "qwen3-coder-480b-a35b-instruct": {
427
+ "command-r-mar-24": {
428
428
  // AA specific benchmarks
429
- codingIndex: 24.6,
430
- mathIndex: 39.3,
429
+ codingIndex: undefined,
430
+ mathIndex: undefined,
431
431
 
432
432
  // Academic benchmarks
433
- mmluPro: 0.788,
434
- gpqa: 0.618,
435
- hle: 0.044,
433
+ mmluPro: 0.338,
434
+ gpqa: 0.284,
435
+ hle: 0.048,
436
436
 
437
437
  // Capabilities
438
438
  contextWindow: 8192,
@@ -440,18 +440,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
440
440
  supportsVision: false,
441
441
 
442
442
  // Metadata
443
- lastUpdated: "2026-06-01",
444
- originalModel: "Qwen3 Coder 480B A35B Instruct",
443
+ lastUpdated: "2026-07-01",
444
+ originalModel: "Command-R (Mar '24)",
445
445
  },
446
- "qwen3.5-27b-reasoning": {
446
+ "apriel-v1.5-15b-thinker": {
447
447
  // AA specific benchmarks
448
- codingIndex: 34.9,
449
- mathIndex: undefined,
448
+ codingIndex: undefined,
449
+ mathIndex: 87.5,
450
450
 
451
451
  // Academic benchmarks
452
- mmluPro: undefined,
453
- gpqa: 0.858,
454
- hle: 0.222,
452
+ mmluPro: 0.773,
453
+ gpqa: 0.713,
454
+ hle: 0.12,
455
455
 
456
456
  // Capabilities
457
457
  contextWindow: 8192,
@@ -459,18 +459,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
459
459
  supportsVision: false,
460
460
 
461
461
  // Metadata
462
- lastUpdated: "2026-06-01",
463
- originalModel: "Qwen3.5 27B (Reasoning)",
462
+ lastUpdated: "2026-07-01",
463
+ originalModel: "Apriel-v1.5-15B-Thinker",
464
464
  },
465
- "qwen3-235b-a22b-reasoning": {
465
+ "jamba-1.5-mini": {
466
466
  // AA specific benchmarks
467
- codingIndex: 17.4,
468
- mathIndex: 82,
467
+ codingIndex: undefined,
468
+ mathIndex: undefined,
469
469
 
470
470
  // Academic benchmarks
471
- mmluPro: 0.828,
472
- gpqa: 0.7,
473
- hle: 0.117,
471
+ mmluPro: 0.371,
472
+ gpqa: 0.302,
473
+ hle: 0.051,
474
474
 
475
475
  // Capabilities
476
476
  contextWindow: 8192,
@@ -478,18 +478,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
478
478
  supportsVision: false,
479
479
 
480
480
  // Metadata
481
- lastUpdated: "2026-06-01",
482
- originalModel: "Qwen3 235B A22B (Reasoning)",
481
+ lastUpdated: "2026-07-01",
482
+ originalModel: "Jamba 1.5 Mini",
483
483
  },
484
- "qwen3-235b-a22b-non-reasoning": {
484
+ "jamba-1.5-large": {
485
485
  // AA specific benchmarks
486
- codingIndex: 14,
487
- mathIndex: 23.7,
486
+ codingIndex: undefined,
487
+ mathIndex: undefined,
488
488
 
489
489
  // Academic benchmarks
490
- mmluPro: 0.762,
491
- gpqa: 0.613,
492
- hle: 0.047,
490
+ mmluPro: 0.572,
491
+ gpqa: 0.427,
492
+ hle: 0.04,
493
493
 
494
494
  // Capabilities
495
495
  contextWindow: 8192,
@@ -497,18 +497,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
497
497
  supportsVision: false,
498
498
 
499
499
  // Metadata
500
- lastUpdated: "2026-06-01",
501
- originalModel: "Qwen3 235B A22B (Non-reasoning)",
500
+ lastUpdated: "2026-07-01",
501
+ originalModel: "Jamba 1.5 Large",
502
502
  },
503
- "qwen3.5-35b-a3b-reasoning": {
503
+ "jamba-1.6-large": {
504
504
  // AA specific benchmarks
505
- codingIndex: 30.3,
505
+ codingIndex: undefined,
506
506
  mathIndex: undefined,
507
507
 
508
508
  // Academic benchmarks
509
- mmluPro: undefined,
510
- gpqa: 0.845,
511
- hle: 0.197,
509
+ mmluPro: 0.565,
510
+ gpqa: 0.387,
511
+ hle: 0.04,
512
512
 
513
513
  // Capabilities
514
514
  contextWindow: 8192,
@@ -516,37 +516,37 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
516
516
  supportsVision: false,
517
517
 
518
518
  // Metadata
519
- lastUpdated: "2026-06-01",
520
- originalModel: "Qwen3.5 35B A3B (Reasoning)",
519
+ lastUpdated: "2026-07-01",
520
+ originalModel: "Jamba 1.6 Large",
521
521
  },
522
- "qwen3-32b-non-reasoning": {
522
+ "jamba-1.6-mini": {
523
523
  // AA specific benchmarks
524
524
  codingIndex: undefined,
525
- mathIndex: 19.7,
525
+ mathIndex: undefined,
526
526
 
527
527
  // Academic benchmarks
528
- mmluPro: 0.727,
529
- gpqa: 0.535,
530
- hle: 0.043,
531
-
528
+ mmluPro: 0.367,
529
+ gpqa: 0.3,
530
+ hle: 0.046,
531
+
532
532
  // Capabilities
533
533
  contextWindow: 8192,
534
534
  supportsReasoning: false,
535
535
  supportsVision: false,
536
536
 
537
537
  // Metadata
538
- lastUpdated: "2026-06-01",
539
- originalModel: "Qwen3 32B (Non-reasoning)",
538
+ lastUpdated: "2026-07-01",
539
+ originalModel: "Jamba 1.6 Mini",
540
540
  },
541
- "qwq-32b-preview": {
541
+ "arctic-instruct": {
542
542
  // AA specific benchmarks
543
543
  codingIndex: undefined,
544
544
  mathIndex: undefined,
545
545
 
546
546
  // Academic benchmarks
547
- mmluPro: 0.648,
548
- gpqa: 0.557,
549
- hle: 0.048,
547
+ mmluPro: undefined,
548
+ gpqa: undefined,
549
+ hle: undefined,
550
550
 
551
551
  // Capabilities
552
552
  contextWindow: 8192,
@@ -554,18 +554,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
554
554
  supportsVision: false,
555
555
 
556
556
  // Metadata
557
- lastUpdated: "2026-06-01",
558
- originalModel: "QwQ 32B-Preview",
557
+ lastUpdated: "2026-07-01",
558
+ originalModel: "Arctic Instruct",
559
559
  },
560
- "qwen3-vl-4b-reasoning": {
560
+ "qwen2.5-max": {
561
561
  // AA specific benchmarks
562
- codingIndex: 6.7,
563
- mathIndex: 25.7,
562
+ codingIndex: undefined,
563
+ mathIndex: undefined,
564
564
 
565
565
  // Academic benchmarks
566
- mmluPro: 0.7,
567
- gpqa: 0.494,
568
- hle: 0.044,
566
+ mmluPro: 0.762,
567
+ gpqa: 0.587,
568
+ hle: 0.045,
569
569
 
570
570
  // Capabilities
571
571
  contextWindow: 8192,
@@ -573,16 +573,73 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
573
573
  supportsVision: false,
574
574
 
575
575
  // Metadata
576
- lastUpdated: "2026-06-01",
577
- originalModel: "Qwen3 VL 4B (Reasoning)",
576
+ lastUpdated: "2026-07-01",
577
+ originalModel: "Qwen2.5 Max",
578
578
  },
579
- "qwen3-vl-4b-instruct": {
579
+ "qwen2.5-instruct-72b": {
580
580
  // AA specific benchmarks
581
- codingIndex: 4.6,
582
- mathIndex: 37,
581
+ codingIndex: undefined,
582
+ mathIndex: 14,
583
583
 
584
584
  // Academic benchmarks
585
- mmluPro: 0.634,
585
+ mmluPro: 0.72,
586
+ gpqa: 0.491,
587
+ hle: 0.042,
588
+
589
+ // Capabilities
590
+ contextWindow: 8192,
591
+ supportsReasoning: false,
592
+ supportsVision: false,
593
+
594
+ // Metadata
595
+ lastUpdated: "2026-07-01",
596
+ originalModel: "Qwen2.5 Instruct 72B",
597
+ },
598
+ "qwen2.5-coder-instruct-32b": {
599
+ // AA specific benchmarks
600
+ codingIndex: undefined,
601
+ mathIndex: undefined,
602
+
603
+ // Academic benchmarks
604
+ mmluPro: 0.635,
605
+ gpqa: 0.417,
606
+ hle: 0.038,
607
+
608
+ // Capabilities
609
+ contextWindow: 8192,
610
+ supportsReasoning: false,
611
+ supportsVision: false,
612
+
613
+ // Metadata
614
+ lastUpdated: "2026-07-01",
615
+ originalModel: "Qwen2.5 Coder Instruct 32B",
616
+ },
617
+ "qwen2.5-turbo": {
618
+ // AA specific benchmarks
619
+ codingIndex: undefined,
620
+ mathIndex: undefined,
621
+
622
+ // Academic benchmarks
623
+ mmluPro: 0.633,
624
+ gpqa: 0.41,
625
+ hle: 0.042,
626
+
627
+ // Capabilities
628
+ contextWindow: 8192,
629
+ supportsReasoning: false,
630
+ supportsVision: false,
631
+
632
+ // Metadata
633
+ lastUpdated: "2026-07-01",
634
+ originalModel: "Qwen2.5 Turbo",
635
+ },
636
+ "qwen2-instruct-72b": {
637
+ // AA specific benchmarks
638
+ codingIndex: undefined,
639
+ mathIndex: undefined,
640
+
641
+ // Academic benchmarks
642
+ mmluPro: 0.622,
586
643
  gpqa: 0.371,
587
644
  hle: 0.037,
588
645
 
@@ -592,18 +649,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
592
649
  supportsVision: false,
593
650
 
594
651
  // Metadata
595
- lastUpdated: "2026-06-01",
596
- originalModel: "Qwen3 VL 4B Instruct",
652
+ lastUpdated: "2026-07-01",
653
+ originalModel: "Qwen2 Instruct 72B",
597
654
  },
598
- "qwen3-30b-a3b-non-reasoning": {
655
+ "qwen3-235b-a22b-2507-instruct": {
599
656
  // AA specific benchmarks
600
- codingIndex: 13.3,
601
- mathIndex: 21.7,
657
+ codingIndex: undefined,
658
+ mathIndex: 71.7,
602
659
 
603
660
  // Academic benchmarks
604
- mmluPro: 0.71,
605
- gpqa: 0.515,
606
- hle: 0.046,
661
+ mmluPro: 0.828,
662
+ gpqa: 0.753,
663
+ hle: 0.106,
607
664
 
608
665
  // Capabilities
609
666
  contextWindow: 8192,
@@ -611,12 +668,202 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
611
668
  supportsVision: false,
612
669
 
613
670
  // Metadata
614
- lastUpdated: "2026-06-01",
615
- originalModel: "Qwen3 30B A3B (Non-reasoning)",
671
+ lastUpdated: "2026-07-01",
672
+ originalModel: "Qwen3 235B A22B 2507 Instruct",
673
+ },
674
+ "qwen3-8b-non-reasoning": {
675
+ // AA specific benchmarks
676
+ codingIndex: undefined,
677
+ mathIndex: 24.3,
678
+
679
+ // Academic benchmarks
680
+ mmluPro: 0.643,
681
+ gpqa: 0.452,
682
+ hle: 0.028,
683
+
684
+ // Capabilities
685
+ contextWindow: 8192,
686
+ supportsReasoning: false,
687
+ supportsVision: false,
688
+
689
+ // Metadata
690
+ lastUpdated: "2026-07-01",
691
+ originalModel: "Qwen3 8B (Non-reasoning)",
692
+ },
693
+ "qwen3-coder-30b-a3b-instruct": {
694
+ // AA specific benchmarks
695
+ codingIndex: undefined,
696
+ mathIndex: 29,
697
+
698
+ // Academic benchmarks
699
+ mmluPro: 0.706,
700
+ gpqa: 0.516,
701
+ hle: 0.04,
702
+
703
+ // Capabilities
704
+ contextWindow: 8192,
705
+ supportsReasoning: false,
706
+ supportsVision: false,
707
+
708
+ // Metadata
709
+ lastUpdated: "2026-07-01",
710
+ originalModel: "Qwen3 Coder 30B A3B Instruct",
711
+ },
712
+ "qwen2.5-instruct-32b": {
713
+ // AA specific benchmarks
714
+ codingIndex: undefined,
715
+ mathIndex: undefined,
716
+
717
+ // Academic benchmarks
718
+ mmluPro: 0.697,
719
+ gpqa: 0.466,
720
+ hle: 0.038,
721
+
722
+ // Capabilities
723
+ contextWindow: 8192,
724
+ supportsReasoning: false,
725
+ supportsVision: false,
726
+
727
+ // Metadata
728
+ lastUpdated: "2026-07-01",
729
+ originalModel: "Qwen2.5 Instruct 32B",
730
+ },
731
+ "qwen3-1.7b-reasoning": {
732
+ // AA specific benchmarks
733
+ codingIndex: undefined,
734
+ mathIndex: 38.7,
735
+
736
+ // Academic benchmarks
737
+ mmluPro: 0.57,
738
+ gpqa: 0.356,
739
+ hle: 0.048,
740
+
741
+ // Capabilities
742
+ contextWindow: 8192,
743
+ supportsReasoning: false,
744
+ supportsVision: false,
745
+
746
+ // Metadata
747
+ lastUpdated: "2026-07-01",
748
+ originalModel: "Qwen3 1.7B (Reasoning)",
749
+ },
750
+ "qwen3-vl-8b-reasoning": {
751
+ // AA specific benchmarks
752
+ codingIndex: undefined,
753
+ mathIndex: 30.7,
754
+
755
+ // Academic benchmarks
756
+ mmluPro: 0.749,
757
+ gpqa: 0.579,
758
+ hle: 0.033,
759
+
760
+ // Capabilities
761
+ contextWindow: 8192,
762
+ supportsReasoning: false,
763
+ supportsVision: false,
764
+
765
+ // Metadata
766
+ lastUpdated: "2026-07-01",
767
+ originalModel: "Qwen3 VL 8B (Reasoning)",
768
+ },
769
+ "qwen3-4b-non-reasoning": {
770
+ // AA specific benchmarks
771
+ codingIndex: undefined,
772
+ mathIndex: undefined,
773
+
774
+ // Academic benchmarks
775
+ mmluPro: 0.586,
776
+ gpqa: 0.398,
777
+ hle: 0.037,
778
+
779
+ // Capabilities
780
+ contextWindow: 8192,
781
+ supportsReasoning: false,
782
+ supportsVision: false,
783
+
784
+ // Metadata
785
+ lastUpdated: "2026-07-01",
786
+ originalModel: "Qwen3 4B (Non-reasoning)",
787
+ },
788
+ "qwen3-4b-reasoning": {
789
+ // AA specific benchmarks
790
+ codingIndex: undefined,
791
+ mathIndex: 22.3,
792
+
793
+ // Academic benchmarks
794
+ mmluPro: 0.696,
795
+ gpqa: 0.522,
796
+ hle: 0.051,
797
+
798
+ // Capabilities
799
+ contextWindow: 8192,
800
+ supportsReasoning: false,
801
+ supportsVision: false,
802
+
803
+ // Metadata
804
+ lastUpdated: "2026-07-01",
805
+ originalModel: "Qwen3 4B (Reasoning)",
806
+ },
807
+ "qwq-32b": {
808
+ // AA specific benchmarks
809
+ codingIndex: undefined,
810
+ mathIndex: 29,
811
+
812
+ // Academic benchmarks
813
+ mmluPro: 0.764,
814
+ gpqa: 0.593,
815
+ hle: 0.082,
816
+
817
+ // Capabilities
818
+ contextWindow: 8192,
819
+ supportsReasoning: false,
820
+ supportsVision: false,
821
+
822
+ // Metadata
823
+ lastUpdated: "2026-07-01",
824
+ originalModel: "QwQ 32B",
825
+ },
826
+ "qwen3-coder-480b-a35b-instruct": {
827
+ // AA specific benchmarks
828
+ codingIndex: undefined,
829
+ mathIndex: 39.3,
830
+
831
+ // Academic benchmarks
832
+ mmluPro: 0.788,
833
+ gpqa: 0.618,
834
+ hle: 0.044,
835
+
836
+ // Capabilities
837
+ contextWindow: 8192,
838
+ supportsReasoning: false,
839
+ supportsVision: false,
840
+
841
+ // Metadata
842
+ lastUpdated: "2026-07-01",
843
+ originalModel: "Qwen3 Coder 480B A35B Instruct",
844
+ },
845
+ "qwen3-235b-a22b-reasoning": {
846
+ // AA specific benchmarks
847
+ codingIndex: undefined,
848
+ mathIndex: 82,
849
+
850
+ // Academic benchmarks
851
+ mmluPro: 0.828,
852
+ gpqa: 0.7,
853
+ hle: 0.117,
854
+
855
+ // Capabilities
856
+ contextWindow: 8192,
857
+ supportsReasoning: false,
858
+ supportsVision: false,
859
+
860
+ // Metadata
861
+ lastUpdated: "2026-07-01",
862
+ originalModel: "Qwen3 235B A22B (Reasoning)",
616
863
  },
617
864
  "qwen3-14b-non-reasoning": {
618
865
  // AA specific benchmarks
619
- codingIndex: 12.4,
866
+ codingIndex: undefined,
620
867
  mathIndex: 58,
621
868
 
622
869
  // Academic benchmarks
@@ -630,18 +877,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
630
877
  supportsVision: false,
631
878
 
632
879
  // Metadata
633
- lastUpdated: "2026-06-01",
880
+ lastUpdated: "2026-07-01",
634
881
  originalModel: "Qwen3 14B (Non-reasoning)",
635
882
  },
636
- "qwen3-32b-reasoning": {
883
+ "qwen3-14b-reasoning": {
637
884
  // AA specific benchmarks
638
885
  codingIndex: 13.8,
639
- mathIndex: 73,
886
+ mathIndex: 55.7,
640
887
 
641
888
  // Academic benchmarks
642
- mmluPro: 0.798,
643
- gpqa: 0.668,
644
- hle: 0.083,
889
+ mmluPro: 0.774,
890
+ gpqa: 0.604,
891
+ hle: 0.043,
645
892
 
646
893
  // Capabilities
647
894
  contextWindow: 8192,
@@ -649,18 +896,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
649
896
  supportsVision: false,
650
897
 
651
898
  // Metadata
652
- lastUpdated: "2026-06-01",
653
- originalModel: "Qwen3 32B (Reasoning)",
899
+ lastUpdated: "2026-07-01",
900
+ originalModel: "Qwen3 14B (Reasoning)",
654
901
  },
655
- "qwen3-vl-8b-reasoning": {
902
+ "qwen3-235b-a22b-non-reasoning": {
656
903
  // AA specific benchmarks
657
- codingIndex: 9.8,
658
- mathIndex: 30.7,
904
+ codingIndex: undefined,
905
+ mathIndex: 23.7,
659
906
 
660
907
  // Academic benchmarks
661
- mmluPro: 0.749,
662
- gpqa: 0.579,
663
- hle: 0.033,
908
+ mmluPro: 0.762,
909
+ gpqa: 0.613,
910
+ hle: 0.047,
664
911
 
665
912
  // Capabilities
666
913
  contextWindow: 8192,
@@ -668,12 +915,50 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
668
915
  supportsVision: false,
669
916
 
670
917
  // Metadata
671
- lastUpdated: "2026-06-01",
672
- originalModel: "Qwen3 VL 8B (Reasoning)",
918
+ lastUpdated: "2026-07-01",
919
+ originalModel: "Qwen3 235B A22B (Non-reasoning)",
920
+ },
921
+ "qwen3-32b-non-reasoning": {
922
+ // AA specific benchmarks
923
+ codingIndex: undefined,
924
+ mathIndex: 19.7,
925
+
926
+ // Academic benchmarks
927
+ mmluPro: 0.727,
928
+ gpqa: 0.535,
929
+ hle: 0.043,
930
+
931
+ // Capabilities
932
+ contextWindow: 8192,
933
+ supportsReasoning: false,
934
+ supportsVision: false,
935
+
936
+ // Metadata
937
+ lastUpdated: "2026-07-01",
938
+ originalModel: "Qwen3 32B (Non-reasoning)",
939
+ },
940
+ "qwen3-32b-reasoning": {
941
+ // AA specific benchmarks
942
+ codingIndex: 15.3,
943
+ mathIndex: 73,
944
+
945
+ // Academic benchmarks
946
+ mmluPro: 0.798,
947
+ gpqa: 0.668,
948
+ hle: 0.083,
949
+
950
+ // Capabilities
951
+ contextWindow: 8192,
952
+ supportsReasoning: false,
953
+ supportsVision: false,
954
+
955
+ // Metadata
956
+ lastUpdated: "2026-07-01",
957
+ originalModel: "Qwen3 32B (Reasoning)",
673
958
  },
674
959
  "qwen3-vl-30b-a3b-reasoning": {
675
960
  // AA specific benchmarks
676
- codingIndex: 13.1,
961
+ codingIndex: undefined,
677
962
  mathIndex: 82.3,
678
963
 
679
964
  // Academic benchmarks
@@ -687,12 +972,31 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
687
972
  supportsVision: false,
688
973
 
689
974
  // Metadata
690
- lastUpdated: "2026-06-01",
975
+ lastUpdated: "2026-07-01",
691
976
  originalModel: "Qwen3 VL 30B A3B (Reasoning)",
692
977
  },
978
+ "qwen3-vl-235b-a22b-instruct": {
979
+ // AA specific benchmarks
980
+ codingIndex: undefined,
981
+ mathIndex: 70.7,
982
+
983
+ // Academic benchmarks
984
+ mmluPro: 0.823,
985
+ gpqa: 0.712,
986
+ hle: 0.063,
987
+
988
+ // Capabilities
989
+ contextWindow: 8192,
990
+ supportsReasoning: false,
991
+ supportsVision: false,
992
+
993
+ // Metadata
994
+ lastUpdated: "2026-07-01",
995
+ originalModel: "Qwen3 VL 235B A22B Instruct",
996
+ },
693
997
  "qwen3-vl-235b-a22b-reasoning": {
694
998
  // AA specific benchmarks
695
- codingIndex: 20.9,
999
+ codingIndex: undefined,
696
1000
  mathIndex: 88.3,
697
1001
 
698
1002
  // Academic benchmarks
@@ -706,18 +1010,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
706
1010
  supportsVision: false,
707
1011
 
708
1012
  // Metadata
709
- lastUpdated: "2026-06-01",
1013
+ lastUpdated: "2026-07-01",
710
1014
  originalModel: "Qwen3 VL 235B A22B (Reasoning)",
711
1015
  },
712
- "qwen3-8b-non-reasoning": {
1016
+ "qwen3-30b-a3b-non-reasoning": {
713
1017
  // AA specific benchmarks
714
- codingIndex: 7.1,
715
- mathIndex: 24.3,
1018
+ codingIndex: undefined,
1019
+ mathIndex: 21.7,
716
1020
 
717
1021
  // Academic benchmarks
718
- mmluPro: 0.643,
719
- gpqa: 0.452,
720
- hle: 0.028,
1022
+ mmluPro: 0.71,
1023
+ gpqa: 0.515,
1024
+ hle: 0.046,
721
1025
 
722
1026
  // Capabilities
723
1027
  contextWindow: 8192,
@@ -725,8 +1029,65 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
725
1029
  supportsVision: false,
726
1030
 
727
1031
  // Metadata
728
- lastUpdated: "2026-06-01",
729
- originalModel: "Qwen3 8B (Non-reasoning)",
1032
+ lastUpdated: "2026-07-01",
1033
+ originalModel: "Qwen3 30B A3B (Non-reasoning)",
1034
+ },
1035
+ "qwen3-235b-a22b-2507-reasoning": {
1036
+ // AA specific benchmarks
1037
+ codingIndex: undefined,
1038
+ mathIndex: 91,
1039
+
1040
+ // Academic benchmarks
1041
+ mmluPro: 0.843,
1042
+ gpqa: 0.79,
1043
+ hle: 0.15,
1044
+
1045
+ // Capabilities
1046
+ contextWindow: 8192,
1047
+ supportsReasoning: false,
1048
+ supportsVision: false,
1049
+
1050
+ // Metadata
1051
+ lastUpdated: "2026-07-01",
1052
+ originalModel: "Qwen3 235B A22B 2507 (Reasoning)",
1053
+ },
1054
+ "qwen3-0.6b-non-reasoning": {
1055
+ // AA specific benchmarks
1056
+ codingIndex: undefined,
1057
+ mathIndex: 10.3,
1058
+
1059
+ // Academic benchmarks
1060
+ mmluPro: 0.231,
1061
+ gpqa: 0.231,
1062
+ hle: 0.052,
1063
+
1064
+ // Capabilities
1065
+ contextWindow: 8192,
1066
+ supportsReasoning: false,
1067
+ supportsVision: false,
1068
+
1069
+ // Metadata
1070
+ lastUpdated: "2026-07-01",
1071
+ originalModel: "Qwen3 0.6B (Non-reasoning)",
1072
+ },
1073
+ "qwen3-4b-2507-reasoning": {
1074
+ // AA specific benchmarks
1075
+ codingIndex: undefined,
1076
+ mathIndex: 82.7,
1077
+
1078
+ // Academic benchmarks
1079
+ mmluPro: 0.743,
1080
+ gpqa: 0.667,
1081
+ hle: 0.059,
1082
+
1083
+ // Capabilities
1084
+ contextWindow: 8192,
1085
+ supportsReasoning: false,
1086
+ supportsVision: false,
1087
+
1088
+ // Metadata
1089
+ lastUpdated: "2026-07-01",
1090
+ originalModel: "Qwen3 4B 2507 (Reasoning)",
730
1091
  },
731
1092
  "qwen3-8b-reasoning": {
732
1093
  // AA specific benchmarks
@@ -744,18 +1105,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
744
1105
  supportsVision: false,
745
1106
 
746
1107
  // Metadata
747
- lastUpdated: "2026-06-01",
1108
+ lastUpdated: "2026-07-01",
748
1109
  originalModel: "Qwen3 8B (Reasoning)",
749
1110
  },
750
- "qwen3-max-thinking-preview": {
1111
+ "qwen3-1.7b-non-reasoning": {
751
1112
  // AA specific benchmarks
752
- codingIndex: 24.5,
753
- mathIndex: 82.3,
1113
+ codingIndex: undefined,
1114
+ mathIndex: 7.3,
754
1115
 
755
1116
  // Academic benchmarks
756
- mmluPro: 0.824,
757
- gpqa: 0.776,
758
- hle: 0.12,
1117
+ mmluPro: 0.411,
1118
+ gpqa: 0.283,
1119
+ hle: 0.052,
759
1120
 
760
1121
  // Capabilities
761
1122
  contextWindow: 8192,
@@ -763,18 +1124,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
763
1124
  supportsVision: false,
764
1125
 
765
1126
  // Metadata
766
- lastUpdated: "2026-06-01",
767
- originalModel: "Qwen3 Max Thinking (Preview)",
1127
+ lastUpdated: "2026-07-01",
1128
+ originalModel: "Qwen3 1.7B (Non-reasoning)",
768
1129
  },
769
- "qwen1.5-chat-110b": {
1130
+ "qwen3.5-35b-a3b-reasoning": {
770
1131
  // AA specific benchmarks
771
1132
  codingIndex: undefined,
772
1133
  mathIndex: undefined,
773
1134
 
774
1135
  // Academic benchmarks
775
1136
  mmluPro: undefined,
776
- gpqa: 0.289,
777
- hle: undefined,
1137
+ gpqa: 0.845,
1138
+ hle: 0.197,
778
1139
 
779
1140
  // Capabilities
780
1141
  contextWindow: 8192,
@@ -782,18 +1143,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
782
1143
  supportsVision: false,
783
1144
 
784
1145
  // Metadata
785
- lastUpdated: "2026-06-01",
786
- originalModel: "Qwen1.5 Chat 110B",
1146
+ lastUpdated: "2026-07-01",
1147
+ originalModel: "Qwen3.5 35B A3B (Reasoning)",
787
1148
  },
788
- "qwen2.5-instruct-32b": {
1149
+ "qwen3.5-27b-reasoning": {
789
1150
  // AA specific benchmarks
790
1151
  codingIndex: undefined,
791
1152
  mathIndex: undefined,
792
1153
 
793
1154
  // Academic benchmarks
794
- mmluPro: 0.697,
795
- gpqa: 0.466,
796
- hle: 0.038,
1155
+ mmluPro: undefined,
1156
+ gpqa: 0.858,
1157
+ hle: 0.222,
797
1158
 
798
1159
  // Capabilities
799
1160
  contextWindow: 8192,
@@ -801,18 +1162,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
801
1162
  supportsVision: false,
802
1163
 
803
1164
  // Metadata
804
- lastUpdated: "2026-06-01",
805
- originalModel: "Qwen2.5 Instruct 32B",
1165
+ lastUpdated: "2026-07-01",
1166
+ originalModel: "Qwen3.5 27B (Reasoning)",
806
1167
  },
807
- "qwen3-vl-30b-a3b-instruct": {
1168
+ "qwen3-vl-8b-instruct": {
808
1169
  // AA specific benchmarks
809
- codingIndex: 14.3,
810
- mathIndex: 72.3,
1170
+ codingIndex: undefined,
1171
+ mathIndex: 27.3,
811
1172
 
812
1173
  // Academic benchmarks
813
- mmluPro: 0.764,
814
- gpqa: 0.695,
815
- hle: 0.064,
1174
+ mmluPro: 0.686,
1175
+ gpqa: 0.427,
1176
+ hle: 0.029,
816
1177
 
817
1178
  // Capabilities
818
1179
  contextWindow: 8192,
@@ -820,18 +1181,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
820
1181
  supportsVision: false,
821
1182
 
822
1183
  // Metadata
823
- lastUpdated: "2026-06-01",
824
- originalModel: "Qwen3 VL 30B A3B Instruct",
1184
+ lastUpdated: "2026-07-01",
1185
+ originalModel: "Qwen3 VL 8B Instruct",
825
1186
  },
826
- "qwen3.5-27b-non-reasoning": {
1187
+ "qwen3-30b-a3b-reasoning": {
827
1188
  // AA specific benchmarks
828
- codingIndex: 33.4,
829
- mathIndex: undefined,
1189
+ codingIndex: undefined,
1190
+ mathIndex: 72.3,
830
1191
 
831
1192
  // Academic benchmarks
832
- mmluPro: undefined,
833
- gpqa: 0.842,
834
- hle: 0.132,
1193
+ mmluPro: 0.777,
1194
+ gpqa: 0.616,
1195
+ hle: 0.066,
835
1196
 
836
1197
  // Capabilities
837
1198
  contextWindow: 8192,
@@ -839,18 +1200,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
839
1200
  supportsVision: false,
840
1201
 
841
1202
  // Metadata
842
- lastUpdated: "2026-06-01",
843
- originalModel: "Qwen3.5 27B (Non-reasoning)",
1203
+ lastUpdated: "2026-07-01",
1204
+ originalModel: "Qwen3 30B A3B (Reasoning)",
844
1205
  },
845
- "qwen3-14b-reasoning": {
1206
+ "qwen3.5-27b-non-reasoning": {
846
1207
  // AA specific benchmarks
847
- codingIndex: 13.1,
848
- mathIndex: 55.7,
1208
+ codingIndex: undefined,
1209
+ mathIndex: undefined,
849
1210
 
850
1211
  // Academic benchmarks
851
- mmluPro: 0.774,
852
- gpqa: 0.604,
853
- hle: 0.043,
1212
+ mmluPro: undefined,
1213
+ gpqa: 0.842,
1214
+ hle: 0.132,
854
1215
 
855
1216
  // Capabilities
856
1217
  contextWindow: 8192,
@@ -858,18 +1219,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
858
1219
  supportsVision: false,
859
1220
 
860
1221
  // Metadata
861
- lastUpdated: "2026-06-01",
862
- originalModel: "Qwen3 14B (Reasoning)",
1222
+ lastUpdated: "2026-07-01",
1223
+ originalModel: "Qwen3.5 27B (Non-reasoning)",
863
1224
  },
864
- "qwen3-30b-a3b-reasoning": {
1225
+ "qwen3-vl-30b-a3b-instruct": {
865
1226
  // AA specific benchmarks
866
- codingIndex: 11,
1227
+ codingIndex: undefined,
867
1228
  mathIndex: 72.3,
868
1229
 
869
1230
  // Academic benchmarks
870
- mmluPro: 0.777,
871
- gpqa: 0.616,
872
- hle: 0.066,
1231
+ mmluPro: 0.764,
1232
+ gpqa: 0.695,
1233
+ hle: 0.064,
873
1234
 
874
1235
  // Capabilities
875
1236
  contextWindow: 8192,
@@ -877,18 +1238,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
877
1238
  supportsVision: false,
878
1239
 
879
1240
  // Metadata
880
- lastUpdated: "2026-06-01",
881
- originalModel: "Qwen3 30B A3B (Reasoning)",
1241
+ lastUpdated: "2026-07-01",
1242
+ originalModel: "Qwen3 VL 30B A3B Instruct",
882
1243
  },
883
- "qwen3-0.6b-non-reasoning": {
1244
+ "qwq-32b-preview": {
884
1245
  // AA specific benchmarks
885
- codingIndex: 1.4,
886
- mathIndex: 10.3,
1246
+ codingIndex: undefined,
1247
+ mathIndex: undefined,
887
1248
 
888
1249
  // Academic benchmarks
889
- mmluPro: 0.231,
890
- gpqa: 0.231,
891
- hle: 0.052,
1250
+ mmluPro: 0.648,
1251
+ gpqa: 0.557,
1252
+ hle: 0.048,
892
1253
 
893
1254
  // Capabilities
894
1255
  contextWindow: 8192,
@@ -896,18 +1257,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
896
1257
  supportsVision: false,
897
1258
 
898
1259
  // Metadata
899
- lastUpdated: "2026-06-01",
900
- originalModel: "Qwen3 0.6B (Non-reasoning)",
1260
+ lastUpdated: "2026-07-01",
1261
+ originalModel: "QwQ 32B-Preview",
901
1262
  },
902
- "qwen3-vl-8b-instruct": {
1263
+ "qwen3-0.6b-reasoning": {
903
1264
  // AA specific benchmarks
904
- codingIndex: 7.3,
905
- mathIndex: 27.3,
1265
+ codingIndex: undefined,
1266
+ mathIndex: 18,
906
1267
 
907
1268
  // Academic benchmarks
908
- mmluPro: 0.686,
909
- gpqa: 0.427,
910
- hle: 0.029,
1269
+ mmluPro: 0.347,
1270
+ gpqa: 0.239,
1271
+ hle: 0.057,
911
1272
 
912
1273
  // Capabilities
913
1274
  contextWindow: 8192,
@@ -915,18 +1276,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
915
1276
  supportsVision: false,
916
1277
 
917
1278
  // Metadata
918
- lastUpdated: "2026-06-01",
919
- originalModel: "Qwen3 VL 8B Instruct",
1279
+ lastUpdated: "2026-07-01",
1280
+ originalModel: "Qwen3 0.6B (Reasoning)",
920
1281
  },
921
- "qwen3-4b-reasoning": {
1282
+ "qwen3-30b-a3b-2507-instruct": {
922
1283
  // AA specific benchmarks
923
1284
  codingIndex: undefined,
924
- mathIndex: 22.3,
1285
+ mathIndex: 66.3,
925
1286
 
926
1287
  // Academic benchmarks
927
- mmluPro: 0.696,
928
- gpqa: 0.522,
929
- hle: 0.051,
1288
+ mmluPro: 0.777,
1289
+ gpqa: 0.659,
1290
+ hle: 0.068,
930
1291
 
931
1292
  // Capabilities
932
1293
  contextWindow: 8192,
@@ -934,18 +1295,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
934
1295
  supportsVision: false,
935
1296
 
936
1297
  // Metadata
937
- lastUpdated: "2026-06-01",
938
- originalModel: "Qwen3 4B (Reasoning)",
1298
+ lastUpdated: "2026-07-01",
1299
+ originalModel: "Qwen3 30B A3B 2507 Instruct",
939
1300
  },
940
- "qwen3-1.7b-non-reasoning": {
1301
+ "qwen3-max-thinking": {
941
1302
  // AA specific benchmarks
942
- codingIndex: 2.3,
943
- mathIndex: 7.3,
1303
+ codingIndex: undefined,
1304
+ mathIndex: undefined,
944
1305
 
945
1306
  // Academic benchmarks
946
- mmluPro: 0.411,
947
- gpqa: 0.283,
948
- hle: 0.052,
1307
+ mmluPro: undefined,
1308
+ gpqa: 0.861,
1309
+ hle: 0.262,
949
1310
 
950
1311
  // Capabilities
951
1312
  contextWindow: 8192,
@@ -953,18 +1314,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
953
1314
  supportsVision: false,
954
1315
 
955
1316
  // Metadata
956
- lastUpdated: "2026-06-01",
957
- originalModel: "Qwen3 1.7B (Non-reasoning)",
1317
+ lastUpdated: "2026-07-01",
1318
+ originalModel: "Qwen3 Max Thinking",
958
1319
  },
959
- "qwen3-vl-32b-reasoning": {
1320
+ "qwen3-max": {
960
1321
  // AA specific benchmarks
961
- codingIndex: 14.5,
962
- mathIndex: 84.7,
1322
+ codingIndex: undefined,
1323
+ mathIndex: 80.7,
963
1324
 
964
1325
  // Academic benchmarks
965
- mmluPro: 0.818,
966
- gpqa: 0.733,
967
- hle: 0.096,
1326
+ mmluPro: 0.841,
1327
+ gpqa: 0.764,
1328
+ hle: 0.111,
968
1329
 
969
1330
  // Capabilities
970
1331
  contextWindow: 8192,
@@ -972,18 +1333,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
972
1333
  supportsVision: false,
973
1334
 
974
1335
  // Metadata
975
- lastUpdated: "2026-06-01",
976
- originalModel: "Qwen3 VL 32B (Reasoning)",
1336
+ lastUpdated: "2026-07-01",
1337
+ originalModel: "Qwen3 Max",
977
1338
  },
978
- "qwen3-30b-a3b-2507-instruct": {
1339
+ "qwen3-vl-4b-instruct": {
979
1340
  // AA specific benchmarks
980
- codingIndex: 14.2,
981
- mathIndex: 66.3,
1341
+ codingIndex: undefined,
1342
+ mathIndex: 37,
982
1343
 
983
1344
  // Academic benchmarks
984
- mmluPro: 0.777,
985
- gpqa: 0.659,
986
- hle: 0.068,
1345
+ mmluPro: 0.634,
1346
+ gpqa: 0.371,
1347
+ hle: 0.037,
987
1348
 
988
1349
  // Capabilities
989
1350
  contextWindow: 8192,
@@ -991,18 +1352,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
991
1352
  supportsVision: false,
992
1353
 
993
1354
  // Metadata
994
- lastUpdated: "2026-06-01",
995
- originalModel: "Qwen3 30B A3B 2507 Instruct",
1355
+ lastUpdated: "2026-07-01",
1356
+ originalModel: "Qwen3 VL 4B Instruct",
996
1357
  },
997
- "qwen3-30b-a3b-2507-reasoning": {
1358
+ "qwen3-vl-4b-reasoning": {
998
1359
  // AA specific benchmarks
999
- codingIndex: 14.6,
1000
- mathIndex: 56.3,
1360
+ codingIndex: undefined,
1361
+ mathIndex: 25.7,
1001
1362
 
1002
1363
  // Academic benchmarks
1003
- mmluPro: 0.805,
1004
- gpqa: 0.707,
1005
- hle: 0.098,
1364
+ mmluPro: 0.7,
1365
+ gpqa: 0.494,
1366
+ hle: 0.044,
1006
1367
 
1007
1368
  // Capabilities
1008
1369
  contextWindow: 8192,
@@ -1010,18 +1371,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1010
1371
  supportsVision: false,
1011
1372
 
1012
1373
  // Metadata
1013
- lastUpdated: "2026-06-01",
1014
- originalModel: "Qwen3 30B A3B 2507 (Reasoning)",
1374
+ lastUpdated: "2026-07-01",
1375
+ originalModel: "Qwen3 VL 4B (Reasoning)",
1015
1376
  },
1016
- "qwen-chat-72b": {
1377
+ "qwen3-4b-2507-instruct": {
1017
1378
  // AA specific benchmarks
1018
1379
  codingIndex: undefined,
1019
- mathIndex: undefined,
1380
+ mathIndex: 52.3,
1020
1381
 
1021
1382
  // Academic benchmarks
1022
- mmluPro: undefined,
1023
- gpqa: undefined,
1024
- hle: undefined,
1383
+ mmluPro: 0.672,
1384
+ gpqa: 0.517,
1385
+ hle: 0.047,
1025
1386
 
1026
1387
  // Capabilities
1027
1388
  contextWindow: 8192,
@@ -1029,12 +1390,12 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1029
1390
  supportsVision: false,
1030
1391
 
1031
1392
  // Metadata
1032
- lastUpdated: "2026-06-01",
1033
- originalModel: "Qwen Chat 72B",
1393
+ lastUpdated: "2026-07-01",
1394
+ originalModel: "Qwen3 4B 2507 Instruct",
1034
1395
  },
1035
1396
  "qwen3-vl-32b-instruct": {
1036
1397
  // AA specific benchmarks
1037
- codingIndex: 15.6,
1398
+ codingIndex: undefined,
1038
1399
  mathIndex: 68.3,
1039
1400
 
1040
1401
  // Academic benchmarks
@@ -1048,18 +1409,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1048
1409
  supportsVision: false,
1049
1410
 
1050
1411
  // Metadata
1051
- lastUpdated: "2026-06-01",
1412
+ lastUpdated: "2026-07-01",
1052
1413
  originalModel: "Qwen3 VL 32B Instruct",
1053
1414
  },
1054
- "qwen3-0.6b-reasoning": {
1415
+ "qwen-chat-72b": {
1055
1416
  // AA specific benchmarks
1056
- codingIndex: 0.9,
1057
- mathIndex: 18,
1417
+ codingIndex: undefined,
1418
+ mathIndex: undefined,
1058
1419
 
1059
1420
  // Academic benchmarks
1060
- mmluPro: 0.347,
1061
- gpqa: 0.239,
1062
- hle: 0.057,
1421
+ mmluPro: undefined,
1422
+ gpqa: undefined,
1423
+ hle: undefined,
1063
1424
 
1064
1425
  // Capabilities
1065
1426
  contextWindow: 8192,
@@ -1067,18 +1428,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1067
1428
  supportsVision: false,
1068
1429
 
1069
1430
  // Metadata
1070
- lastUpdated: "2026-06-01",
1071
- originalModel: "Qwen3 0.6B (Reasoning)",
1431
+ lastUpdated: "2026-07-01",
1432
+ originalModel: "Qwen Chat 72B",
1072
1433
  },
1073
- "qwen3-1.7b-reasoning": {
1434
+ "qwen3-max-preview": {
1074
1435
  // AA specific benchmarks
1075
- codingIndex: 1.4,
1076
- mathIndex: 38.7,
1436
+ codingIndex: undefined,
1437
+ mathIndex: 75,
1077
1438
 
1078
1439
  // Academic benchmarks
1079
- mmluPro: 0.57,
1080
- gpqa: 0.356,
1081
- hle: 0.048,
1440
+ mmluPro: 0.838,
1441
+ gpqa: 0.764,
1442
+ hle: 0.093,
1082
1443
 
1083
1444
  // Capabilities
1084
1445
  contextWindow: 8192,
@@ -1086,18 +1447,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1086
1447
  supportsVision: false,
1087
1448
 
1088
1449
  // Metadata
1089
- lastUpdated: "2026-06-01",
1090
- originalModel: "Qwen3 1.7B (Reasoning)",
1450
+ lastUpdated: "2026-07-01",
1451
+ originalModel: "Qwen3 Max (Preview)",
1091
1452
  },
1092
- "qwen3-max-preview": {
1453
+ "qwen2.5-coder-instruct-7b": {
1093
1454
  // AA specific benchmarks
1094
- codingIndex: 25.5,
1095
- mathIndex: 75,
1455
+ codingIndex: undefined,
1456
+ mathIndex: undefined,
1096
1457
 
1097
1458
  // Academic benchmarks
1098
- mmluPro: 0.838,
1099
- gpqa: 0.764,
1100
- hle: 0.093,
1459
+ mmluPro: 0.473,
1460
+ gpqa: 0.339,
1461
+ hle: 0.048,
1101
1462
 
1102
1463
  // Capabilities
1103
1464
  contextWindow: 8192,
@@ -1105,18 +1466,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1105
1466
  supportsVision: false,
1106
1467
 
1107
1468
  // Metadata
1108
- lastUpdated: "2026-06-01",
1109
- originalModel: "Qwen3 Max (Preview)",
1469
+ lastUpdated: "2026-07-01",
1470
+ originalModel: "Qwen2.5 Coder Instruct 7B ",
1110
1471
  },
1111
- "qwen3.6-max-preview": {
1472
+ "qwen3-max-thinking-preview": {
1112
1473
  // AA specific benchmarks
1113
- codingIndex: 44.9,
1114
- mathIndex: undefined,
1474
+ codingIndex: undefined,
1475
+ mathIndex: 82.3,
1115
1476
 
1116
1477
  // Academic benchmarks
1117
- mmluPro: undefined,
1118
- gpqa: 0.888,
1119
- hle: 0.289,
1478
+ mmluPro: 0.824,
1479
+ gpqa: 0.776,
1480
+ hle: 0.12,
1120
1481
 
1121
1482
  // Capabilities
1122
1483
  contextWindow: 8192,
@@ -1124,18 +1485,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1124
1485
  supportsVision: false,
1125
1486
 
1126
1487
  // Metadata
1127
- lastUpdated: "2026-06-01",
1128
- originalModel: "Qwen3.6 Max Preview",
1488
+ lastUpdated: "2026-07-01",
1489
+ originalModel: "Qwen3 Max Thinking (Preview)",
1129
1490
  },
1130
- "qwen3-max": {
1491
+ "qwen3-30b-a3b-2507-reasoning": {
1131
1492
  // AA specific benchmarks
1132
- codingIndex: 26.4,
1133
- mathIndex: 80.7,
1493
+ codingIndex: undefined,
1494
+ mathIndex: 56.3,
1134
1495
 
1135
1496
  // Academic benchmarks
1136
- mmluPro: 0.841,
1137
- gpqa: 0.764,
1138
- hle: 0.111,
1497
+ mmluPro: 0.805,
1498
+ gpqa: 0.707,
1499
+ hle: 0.098,
1139
1500
 
1140
1501
  // Capabilities
1141
1502
  contextWindow: 8192,
@@ -1143,18 +1504,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1143
1504
  supportsVision: false,
1144
1505
 
1145
1506
  // Metadata
1146
- lastUpdated: "2026-06-01",
1147
- originalModel: "Qwen3 Max",
1507
+ lastUpdated: "2026-07-01",
1508
+ originalModel: "Qwen3 30B A3B 2507 (Reasoning)",
1148
1509
  },
1149
- "qwen3-max-thinking": {
1510
+ "qwen3.6-max-preview": {
1150
1511
  // AA specific benchmarks
1151
- codingIndex: 30.5,
1512
+ codingIndex: undefined,
1152
1513
  mathIndex: undefined,
1153
1514
 
1154
1515
  // Academic benchmarks
1155
1516
  mmluPro: undefined,
1156
- gpqa: 0.861,
1157
- hle: 0.262,
1517
+ gpqa: 0.888,
1518
+ hle: 0.289,
1158
1519
 
1159
1520
  // Capabilities
1160
1521
  contextWindow: 8192,
@@ -1162,18 +1523,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1162
1523
  supportsVision: false,
1163
1524
 
1164
1525
  // Metadata
1165
- lastUpdated: "2026-06-01",
1166
- originalModel: "Qwen3 Max Thinking",
1526
+ lastUpdated: "2026-07-01",
1527
+ originalModel: "Qwen3.6 Max Preview",
1167
1528
  },
1168
- "qwen3-4b-2507-reasoning": {
1529
+ "qwen1.5-chat-110b": {
1169
1530
  // AA specific benchmarks
1170
- codingIndex: 9.5,
1171
- mathIndex: 82.7,
1531
+ codingIndex: undefined,
1532
+ mathIndex: undefined,
1172
1533
 
1173
1534
  // Academic benchmarks
1174
- mmluPro: 0.743,
1175
- gpqa: 0.667,
1176
- hle: 0.059,
1535
+ mmluPro: undefined,
1536
+ gpqa: 0.289,
1537
+ hle: undefined,
1177
1538
 
1178
1539
  // Capabilities
1179
1540
  contextWindow: 8192,
@@ -1181,18 +1542,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1181
1542
  supportsVision: false,
1182
1543
 
1183
1544
  // Metadata
1184
- lastUpdated: "2026-06-01",
1185
- originalModel: "Qwen3 4B 2507 (Reasoning)",
1545
+ lastUpdated: "2026-07-01",
1546
+ originalModel: "Qwen1.5 Chat 110B",
1186
1547
  },
1187
- "qwen3-4b-2507-instruct": {
1548
+ "qwen3-vl-32b-reasoning": {
1188
1549
  // AA specific benchmarks
1189
- codingIndex: 9,
1190
- mathIndex: 52.3,
1550
+ codingIndex: undefined,
1551
+ mathIndex: 84.7,
1191
1552
 
1192
1553
  // Academic benchmarks
1193
- mmluPro: 0.672,
1194
- gpqa: 0.517,
1195
- hle: 0.047,
1554
+ mmluPro: 0.818,
1555
+ gpqa: 0.733,
1556
+ hle: 0.096,
1196
1557
 
1197
1558
  // Capabilities
1198
1559
  contextWindow: 8192,
@@ -1200,18 +1561,18 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1200
1561
  supportsVision: false,
1201
1562
 
1202
1563
  // Metadata
1203
- lastUpdated: "2026-06-01",
1204
- originalModel: "Qwen3 4B 2507 Instruct",
1564
+ lastUpdated: "2026-07-01",
1565
+ originalModel: "Qwen3 VL 32B (Reasoning)",
1205
1566
  },
1206
- "qwen2.5-coder-instruct-7b": {
1567
+ "ring-1t": {
1207
1568
  // AA specific benchmarks
1208
1569
  codingIndex: undefined,
1209
- mathIndex: undefined,
1570
+ mathIndex: 89.3,
1210
1571
 
1211
1572
  // Academic benchmarks
1212
- mmluPro: 0.473,
1213
- gpqa: 0.339,
1214
- hle: 0.048,
1573
+ mmluPro: 0.806,
1574
+ gpqa: 0.774,
1575
+ hle: 0.102,
1215
1576
 
1216
1577
  // Capabilities
1217
1578
  contextWindow: 8192,
@@ -1219,12 +1580,12 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1219
1580
  supportsVision: false,
1220
1581
 
1221
1582
  // Metadata
1222
- lastUpdated: "2026-06-01",
1223
- originalModel: "Qwen2.5 Coder Instruct 7B ",
1583
+ lastUpdated: "2026-07-01",
1584
+ originalModel: "Ring-1T",
1224
1585
  },
1225
1586
  "ling-1t": {
1226
1587
  // AA specific benchmarks
1227
- codingIndex: 18.8,
1588
+ codingIndex: undefined,
1228
1589
  mathIndex: 71.3,
1229
1590
 
1230
1591
  // Academic benchmarks
@@ -1238,12 +1599,12 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1238
1599
  supportsVision: false,
1239
1600
 
1240
1601
  // Metadata
1241
- lastUpdated: "2026-06-01",
1602
+ lastUpdated: "2026-07-01",
1242
1603
  originalModel: "Ling-1T",
1243
1604
  },
1244
1605
  "ling-flash-2.0": {
1245
1606
  // AA specific benchmarks
1246
- codingIndex: 16.7,
1607
+ codingIndex: undefined,
1247
1608
  mathIndex: 65.3,
1248
1609
 
1249
1610
  // Academic benchmarks
@@ -1257,31 +1618,12 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1257
1618
  supportsVision: false,
1258
1619
 
1259
1620
  // Metadata
1260
- lastUpdated: "2026-06-01",
1621
+ lastUpdated: "2026-07-01",
1261
1622
  originalModel: "Ling-flash-2.0",
1262
1623
  },
1263
- "ring-1t": {
1264
- // AA specific benchmarks
1265
- codingIndex: 16.8,
1266
- mathIndex: 89.3,
1267
-
1268
- // Academic benchmarks
1269
- mmluPro: 0.806,
1270
- gpqa: 0.774,
1271
- hle: 0.102,
1272
-
1273
- // Capabilities
1274
- contextWindow: 8192,
1275
- supportsReasoning: false,
1276
- supportsVision: false,
1277
-
1278
- // Metadata
1279
- lastUpdated: "2026-06-01",
1280
- originalModel: "Ring-1T",
1281
- },
1282
1624
  "seed-oss-36b-instruct": {
1283
1625
  // AA specific benchmarks
1284
- codingIndex: 16.7,
1626
+ codingIndex: undefined,
1285
1627
  mathIndex: 84.7,
1286
1628
 
1287
1629
  // Academic benchmarks
@@ -1295,7 +1637,7 @@ export const BENCHMARKS_CHUNK_5: Record<string, HardcodedBenchmark> = {
1295
1637
  supportsVision: false,
1296
1638
 
1297
1639
  // Metadata
1298
- lastUpdated: "2026-06-01",
1640
+ lastUpdated: "2026-07-01",
1299
1641
  originalModel: "Seed-OSS-36B-Instruct",
1300
1642
  },
1301
1643
  };