Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d261Policy30d64, 10 pendingSources1,015
Pro Mode — raw scores, one benchmark under the lensofficial card harness
The field · 62 measured
3.33100

62 models measured, most on the official card harness. Claude Opus 4.8 tops the board at 100.

62 measured·6 new this week·Lifecycle Updated Oct 9, 2026Score basis
#◆Vendor$/M in$/M out
Leader — ignites whiteField — fill cools with scoreBars show 3.33–100 · ◆ solid = first-party or better · outlined = arm's-length
01Claude Opus 4.8+3 altsAnthropic1003P$5.0/M$25.0/M$0.150
02GPT-5.5+3 altsOpenAI1003P$5.0/M$30.0/M$0.175
03GPT-5.6 SolOpenAI99.93P$4.0/M$20.0/M$0.120
04GLM-5.2+6 altsZ.ai99.2$1.4/M$4.4/M$0.029
05GPT-5.4+5 altsOpenAI99.173P$2.5/M$15.0/M$0.088
06Gemini 3.1 Pro+9 altsGoogle98.333P$2.0/M$12.0/M$0.071
07GPT-5.2+1 altOpenAI98.333P$1.8/M$14.0/M$0.080
08GPT-5.6 Luna+1 altOpenAI97.6$0.20/M$1.2/M$0.007
09Inkling+3 altsThinking Machines97.1$0.95/M$4.0/M$0.026
10Qwen3.7 Max+1 altAlibaba97$2.5/M$7.5/M$0.052
11Namazusakana96.67$0.95/M$4.0/M$0.026
12Gemini 3.6 FlashGoogle96.673P$0.75/M$3.8/M$0.023
13DeepSeek-V4-Pro+5 altsDeepSeek96.673P$0.43/M$0.87/M$0.007
14Kimi K3Moonshot96.673P$0.58/M$12.3/M$0.067
15Gemini 3 Flash Preview+1 altGoogle96.673P$0.50/M$3.0/M$0.018
16Step 3.5 Flash+2 altsStepFun96.67$0.10/M$0.30/M$0.002
17Claude Opus 4.6+5 altsAnthropic96.673P$5.0/M$25.0/M$0.155
18Kimi K2.6+4 altsMoonshot96.4$0.95/M$4.0/M$0.026
19Claude Opus 4.7+1 altAnthropic95.833P$5.0/M$25.0/M$0.157
20DeepSeek-V4-Flash+3 altsDeepSeek95.833P$0.44/M$1.3/M$0.009
21Kimi K2.5+6 altsMoonshot95.833P$0.45/M$2.3/M$0.014
22Inkling-Small+2 altsThinking Machines95.5$0.30/M$1.2/M$0.008
23GLM-5+4 altsZ.ai95.4$1.0/M$3.2/M$0.022
24GLM-5.1+6 altsZ.ai95.3$1.4/M$4.4/M$0.030
25Qwen3.6 Plus+3 altsAlibaba95.3$0.50/M$3.0/M$0.018
26Solar Pro4Upstage95.3$0.30/M$1.2/M$0.008
27Claude Opus 4.5Anthropic95.13P$5.0/M$25.0/M$0.158
28Step 3.7 Flash+1 altStepFun953P$0.20/M$1.1/M$0.007
29Gemini 3.5 Flash+1 altGoogle953P$1.5/M$9.0/M$0.055
30Muse GlimmerMeta94.7$0.33/M$1.4/M$0.009
31MAI-Thinking-1+1 altMicrosoft94.5———
32Seed 2.0ByteDance94.2$0.50/M$3.0/M$0.019
33Nemotron 3 Ultra 550B A55B+2 altsNVIDIA94.2$0.60/M$2.5/M$0.016
34DeepSeek-V3.2+4 altsDeepSeek94.173P$0.28/M$0.42/M$0.004
35Grok 4.1 Fast+1 altSpaceXAI94.173P$0.20/M$0.50/M$0.004
36Qwen3.6 27B+1 altAlibaba94.1$0.60/M$3.6/M$0.022
37MiMo V2.5+1 altXiaomi93.6$0.14/M$0.28/M$0.002
38Qwen3.5 35B A3B+3 altsAlibaba93.333P$0.25/M$2.0/M$0.012
39Qwen3 Max (Reasoning)Alibaba93.3$1.2/M$6.0/M$0.039
40Ling-3.0-flashInclusionAI93.2$0.07/M$0.22/M$0.002
41Qwen3.6 35B A3B+2 altsAlibaba92.7$0.38/M$2.3/M$0.014
42EXAONE 4.5 33BLG AI92.6———
43MAI-Code-1-FlashMicrosoft92.5———
44Qwen3.5 9B+1 altAlibaba92.53P$0.14/M$0.20/M$0.002
45Nemotron 3 Super 120B A12B+1 altNVIDIA91.673P$0.30/M$0.90/M$0.007
46Gemini 3 Pro+2 altsGoogle91.673P$2.0/M$12.0/M$0.076
47Qwen3.5 27B+3 altsAlibaba91.673P$0.30/M$2.4/M$0.015
48Qwen3.5 397B A17B+5 altsAlibaba91.3$0.60/M$3.6/M$0.023
49MiniMax M2.7NEW+4 altsMiniMax89.8$0.21/M$0.84/M$0.006
50Qwen3.5 4B+2 altsAlibaba89.673P$0.03/M$0.15/M$0.001
51Gemma 4 31BNEW+29 altsGoogle89.2$0.17/M$0.40/M$0.003
52Qwen3 30B A3B 2507 Thinking+3 altsAlibaba88.333P$0.20/M$2.4/M$0.015
53Gemma 4 26B A4BNEW+33 altsGoogle88.3$0.07/M$0.34/M$0.002
54QED-Nano+1 altLM-Provers82.53P———
55Qwen3 4B 2507 Thinking+2 altsAlibaba82.53P$0.01/M$0.03/M$0.000
56Gemini 3.5 Flash Lite+1 altGoogle82.2$0.30/M$2.5/M$0.017
57Gemma 4 12B+25 altsGoogle77.5$0.10/M$0.30/M$0.003
58LFM2.5-8B-A1BLiquid AI503P———
59Gemma 4 E4BNEW+33 altsGoogle42.5$0.02/M$0.10/M$0.001
60Gemma 4 E2BNEW+38 altsGoogle37.5———
61Gemma 3 27BNEW+27 altsGoogle20.8$0.08/M$0.16/M$0.006
62Granite 4.0 H TinyIBM3.333P———
◆ solid = independently verified · aggregator attested · vendor attributed — outlined = cross-reference · source-attributed. Dimmed ◆ = confidence inferred. “+N alts” = alternate disclosures in the drawer with why-demoted. Prices per million tokens; $/point = input price ÷ score. Every row opens its model dossier.

Verification: 62 models scored · 26 independently verified · 9 vendor cross-reference · 27 vendor-reported · 0 with source disagreement. How these tiers are assigned