Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models739Deals30d256Policy30d64, 14 pendingSources1,010
Pro Mode — raw scores, one benchmark under the lensofficial card harness
The field · 31 measured
27.381.7

31 models measured, most on the official card harness. Qwen3.7 Plus Preview tops the board at 81.7.

31 measured·Lifecycle Updated Oct 8, 2026Score basis
#◆Vendor$/M in$/M out
Leader — ignites whiteField — fill cools with scoreBars show 27.3–81.7 · ◆ solid = first-party or better · outlined = arm's-length
01Qwen3.7 Plus PreviewAlibaba81.7$0.40/M$1.6/M$0.012
02GLM 5V TurboZ.ai78.2$1.2/M$4.0/M$0.033
03Seed 2.1 Pro Preview+1 altByteDance74.5———
04Muse SparkMeta71.3———
05Kimi K2.5+1 altMoonshot71.2$0.45/M$2.3/M$0.019
06Seed2.1ByteDance71.1———
07Gemini 3.1 ProGoogle69.9$2.0/M$12.0/M$0.100
08Claude Opus 4.5Anthropic69.73P$5.0/M$25.0/M$0.215
09Gemini 3 ProGoogle69.73P$2.0/M$12.0/M$0.100
10Qwen3.6 PlusAlibaba67.3$0.50/M$3.0/M$0.026
11Seed 1.8ByteDance65.4$0.25/M$2.0/M$0.017
12Qwen3.5 122B A10BAlibaba61.7$0.40/M$3.2/M$0.029
13Qwen3 VL 235B A22B Reasoning+1 altAlibaba61.3$0.40/M$4.0/M$0.036
14Qwen3.6 35B A3B+1 altAlibaba58.9$0.38/M$2.3/M$0.022
15GPT-5.5OpenAI58.6$5.0/M$30.0/M$0.299
16Qwen3.5 35B A3B+1 altAlibaba58.3$0.25/M$2.0/M$0.019
17Claude Sonnet 4.5Anthropic57.6$3.0/M$15.0/M$0.156
18Claude Opus 4.7Anthropic56.5$5.0/M$25.0/M$0.265
19Qwen3.6 27BAlibaba56.1$0.60/M$3.6/M$0.037
20Qwen3.5 27B+1 altAlibaba56$0.30/M$2.4/M$0.024
21GPT-5.2OpenAI55.83P$1.8/M$14.0/M$0.141
22Gemma 4 31BGoogle52.9$0.17/M$0.40/M$0.005
23Gemma 4 26B A4BGoogle52.2$0.07/M$0.34/M$0.004
24Qwen3.5 4B+1 altAlibaba40.7$0.03/M$0.15/M$0.002
25LFM2.5-VL-3B+1 altLiquid AI35.4———
26Qwen3.5 2BAlibaba35.2———
27InternVL3_5-4BOpenGVLab33.7———
28LFM2-VL-3BLiquid AI33———
29InternVL3_5-2BOpenGVLab30.5———
30Gemma 4 E4BGoogle30.4$0.02/M$0.10/M$0.002
31Gemma 4 E2BGoogle27.3———
◆ 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: 31 models scored · 0 independently verified · 9 vendor cross-reference · 22 vendor-reported · 1 with source disagreement. How these tiers are assigned