Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models762Deals30d246Policy30d59, 11 pendingSources1,061
Pro Mode — raw scores, one benchmark under the lensofficial card harness
The field · 24 measured
27.286.67

24 models measured, most on the official card harness. Nemotron 3 Super 120B A12B tops the board at 86.67.

24 measured·24 new this week·Lifecycle Updated Oct 17, 2026Score basis
#◆Vendor$/M in$/M out
Leader — ignites whiteField — fill cools with scoreBars show 27.2–86.67 · ◆ solid = first-party or better · outlined = arm's-length
01Nemotron 3 Super 120B A12BNEWNVIDIA86.67$0.30/M$0.90/M$0.007
02Qwen3.8 Flash NextNEWAlibaba86.23P$0.15/M$0.47/M$0.004
03NVIDIA Nemotron 3 Nano 30B A3BNEWNVIDIA86.2$0.05/M$0.20/M$0.001
04Qwen3.7 MaxNEWAlibaba85.8$2.5/M$7.5/M$0.058
05DeepSeek-V4-FlashNEWDeepSeek85.53P$0.01/M$0.30/M$0.002
06NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16NEWNVIDIA84.83P———
07Qwen3.7 Plus PreviewNEWAlibaba84.6$0.40/M$1.6/M$0.012
08Inkling-SmallNEWThinking Machines84.63P$0.30/M$1.2/M$0.009
09Qwen3.6 PlusNEWAlibaba84.3$0.50/M$3.0/M$0.021
10Nemotron 3 Ultra 550B A55BNEW+1 altNVIDIA83.7$0.60/M$2.7/M$0.020
11Command A+NEWCohere81$0.30/M$1.5/M$0.011
12Qwen3.5 397B A17BNEWAlibaba78.9$0.60/M$3.6/M$0.027
13Qwen3.5 122B A10BNEW+1 altAlibaba78.3$0.40/M$3.2/M$0.023
14Qwen3 Max (Reasoning)NEWAlibaba77.6$1.2/M$6.0/M$0.046
15Qwen3.5 27BNEWAlibaba77.6$0.30/M$2.4/M$0.017
16Qwen3.5 35B A3BNEWAlibaba76.3$0.25/M$2.0/M$0.015
17Qwen3.5 9BNEWAlibaba72.6$0.14/M$0.20/M$0.002
18Qwen3.5 4BNEWAlibaba66.6$0.03/M$0.15/M$0.001
19Gemma 3 27BNEWGoogle53.4$0.08/M$0.16/M$0.002
20Gemma 3 12BNEWGoogle51.6$0.05/M$0.15/M$0.002
21Gemma 3 4BNEWGoogle46.8$0.05/M$0.10/M$0.002
22Qwen3.5 2BNEWAlibaba45.8———
23Gemma 3 1B InstructNEWGoogle35.9———
24Qwen3.5 0.8BNEWAlibaba27.2———
◆ 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: 24 models scored · 0 independently verified · 3 vendor cross-reference · 21 vendor-reported · 0 with source disagreement. How these tiers are assigned