Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d261Policy30d65, 14 pendingSources1,013
Pro Mode — raw scores, one benchmark under the lensofficial card harness
The field · 71 measured
39.793.2

71 models measured, most on the official card harness. Claude Mythos Preview tops the board at 93.2.

71 measured·Lifecycle Updated Oct 8, 2026Score basis
#◆Vendor$/M in$/M out
Leader — ignites whiteField — fill cools with scoreBars show 39.7–93.2 · ◆ solid = first-party or better · outlined = arm's-length
01Claude Mythos Preview+1 altAnthropic93.2———
02Kimi K3Moonshot91.3$0.58/M$12.3/M$0.071
03Claude Opus 4.7+4 altsAnthropic91$5.0/M$25.0/M$0.165
04Qwen3.8 Flash Next+1 altAlibaba90.6$0.15/M$0.47/M$0.003
05Qwen3.8 27B+1 altAlibaba90.2$0.50/M$3.0/M$0.019
06Claude Opus 4.8+3 altsAnthropic89.9$5.0/M$25.0/M$0.167
07GLM 5.3 FlashZ.ai89.4$0.15/M$0.50/M$0.004
08Gemini 3.6 Flash+1 altGoogle89.4$0.75/M$3.8/M$0.025
09Claude Mythos 5Anthropic88.9$10.0/M$50.0/M$0.337
10Gemini 3.7 Flash+2 altsGoogle88.7$0.75/M$3.8/M$0.025
11Muse Spark 1.1Meta88.4$1.3/M$4.3/M$0.031
12Claude Sonnet 5+2 altsAnthropic88.3$2.0/M$10.0/M$0.068
13Kimi K2.6+1 altMoonshot86.7$0.95/M$4.0/M$0.029
14Muse SparkMeta86.4———
15Seed 2.1 Pro PreviewByteDance86.4———
16Gemini 3.8 Flash+2 altsGoogle86.2$0.75/M$3.8/M$0.026
17Qwen3.7 Plus Preview+1 altAlibaba85.9$0.40/M$1.6/M$0.012
18GPT-5.6 TerraOpenAI85.9$2.0/M$12.0/M$0.081
19GPT-5.6 SolOpenAI85.8$4.0/M$20.0/M$0.140
20Gemini 3.5 Flash+2 altsGoogle84.2$1.5/M$9.0/M$0.062
21Claude Opus 5Anthropic83.7$5.0/M$25.0/M$0.179
22Gemini 3.1 Pro+2 altsGoogle83.3$2.0/M$12.0/M$0.084
23GPT-5.4OpenAI82.83P$2.5/M$15.0/M$0.106
24GPT-5.6 LunaOpenAI82.7$0.20/M$1.2/M$0.008
25GPT-5.2+1 altOpenAI82.1$1.8/M$14.0/M$0.096
26Grok4.5SpaceXAI81.6$2.0/M$6.0/M$0.049
27GPT-5.5 InstantOpenAI81.6$5.0/M$30.0/M$0.214
28Qwen3.6 PlusAlibaba81.5$0.50/M$3.0/M$0.021
29Gemini 3 Pro+1 altGoogle81.4$2.0/M$12.0/M$0.086
30GPT-5OpenAI81.1$1.3/M$10.0/M$0.069
31MiMo V2.5Xiaomi81$0.14/M$0.28/M$0.003
32Gemini 3 Flash Preview+1 altGoogle80.3$0.50/M$3.0/M$0.022
33Qwen3.5 27B+1 altAlibaba79.5$0.30/M$2.4/M$0.017
34Muse GlimmerMeta78.8$0.33/M$1.4/M$0.011
35O3OpenAI78.6$2.0/M$8.0/M$0.064
36Qwen3.6 27BAlibaba78.4$0.60/M$3.6/M$0.027
37InklingThinking Machines78.1$0.95/M$4.0/M$0.032
38Qwen3.6 35B A3B+1 altAlibaba78$0.38/M$2.3/M$0.017
39Qwen3.5 35B A3B+1 altAlibaba77.5$0.25/M$2.0/M$0.015
40Kimi K2.5+2 altsMoonshot77.5$0.45/M$2.3/M$0.017
41Inkling-SmallThinking Machines77.4$0.30/M$1.2/M$0.010
42Claude Opus 4.6+4 altsAnthropic77.4$5.0/M$25.0/M$0.194
43Qwen3.5 122B A10BAlibaba77.2$0.40/M$3.2/M$0.023
44Gemini 3.5 Flash Lite+1 altGoogle76.5$0.30/M$2.5/M$0.018
45Gemini 3.1 Flash Lite PreviewGoogle73.2$0.25/M$1.5/M$0.012
46O4 MiniOpenAI72$1.1/M$4.4/M$0.038
47Claude Sonnet 4.6+1 altAnthropic71.6$3.0/M$15.0/M$0.126
48Seed 1.8ByteDance71.4$0.25/M$2.0/M$0.016
49Gemma 4 26B A4BGoogle69$0.07/M$0.34/M$0.003
50Gemma 4 31BGoogle67.9$0.17/M$0.40/M$0.004
51Claude Opus 4.5Anthropic67.23P$5.0/M$25.0/M$0.223
52Claude Sonnet 4.5Anthropic67.2$3.0/M$15.0/M$0.134
53Qwen3 VL 235B A22B Reasoning+1 altAlibaba66.1$0.40/M$4.0/M$0.033
54Qwen3 VL 32B ReasoningAlibaba65.2$0.16/M$0.64/M$0.006
55Nemotron 3 Nano OmniNVIDIA63.63P$0.30/M$0.90/M$0.009
56Qwen3 VL 32B InstructAlibaba62.8$0.16/M$0.64/M$0.006
57Qwen3 VL 235B A22B InstructAlibaba62.1$0.40/M$1.6/M$0.016
58GPT-4oOpenAI58.8$5.0/M$15.0/M$0.170
59GPT-4.1 miniOpenAI56.8$0.40/M$1.6/M$0.018
60GPT-4.1OpenAI56.7$2.0/M$8.0/M$0.088
61Qwen3 VL 30B A3B ReasoningAlibaba56.6$0.20/M$2.4/M$0.023
62GPT-4.5 PreviewOpenAI55.4———
63Qwen3 VL Thinking (8B)Alibaba53$0.18/M$2.1/M$0.022
64Command A+Cohere52.7$0.30/M$1.5/M$0.017
65Qwen3 VL 4B (Reasoning)+1 altAlibaba50.3———
66Qwen3 VL 30B A3B InstructAlibaba48.9$0.20/M$0.80/M$0.010
67Qwen3 VL 8B InstructAlibaba46.4$0.18/M$0.70/M$0.009
68NVIDIA Nemotron Nano 12B v2 VLNVIDIA41.33P$0.20/M$0.60/M$0.010
69Qwen3 VL 2BAlibaba41.33P———
70GPT-4.1 nanoOpenAI40.5$0.10/M$0.40/M$0.006
71Qwen3 VL 4B InstructAlibaba39.7———
◆ 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: 71 models scored · 0 independently verified · 11 vendor cross-reference · 60 vendor-reported · 7 with source disagreement. How these tiers are assigned