Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models727Deals30d251Policy30d58, 11 pendingSources984
Pro Mode — raw scores, one benchmark under the lensthird-party eval harness
The field · 69 measured
5.9693.68

69 models measured, most on the third-party eval harness. GPT-6 Astra tops the board at 93.68.

69 measured·Lifecycle Updated Sep 29, 2026Basis
#◆Vendor$/M in$/M out
Leader — ignites whiteField — fill cools with scoreBars show 5.96–93.68 · ◆ solid = first-party or better · outlined = arm's-length
01GPT-6 AstraOpenAI93.68$10.0/M$50.0/M$0.320
02Claude Fable 5.1Anthropic90.18$10.0/M$50.0/M$0.333
03GPT-5.6 SolOpenAI89.12$4.0/M$20.0/M$0.135
04GPT-5.5 ProOpenAI87.72$30.0/M$180.0/M$1.197
05Claude Fable 5Anthropic87.02$10.0/M$50.0/M$0.345
06GPT-5.6 TerraOpenAI85.96$2.0/M$12.0/M$0.081
07Claude Opus 5Anthropic85.61$5.0/M$25.0/M$0.175
08GPT-5.5OpenAI85.26$5.0/M$30.0/M$0.205
09GPT-5.4 ProOpenAI82.46$30.0/M$180.0/M$1.273
10GPT-5.6 Luna+2 altsOpenAI82.11$0.20/M$1.2/M$0.009
11Claude Opus 4.8Anthropic80$5.0/M$25.0/M$0.188
12GPT-5.4OpenAI78.6$2.5/M$15.0/M$0.111
13Qwen3.8 Max Preview+1 altAlibaba74.74$3.3/M$9.9/M$0.088
14Muse Spark 1.3+1 altMeta74.04$1.3/M$4.3/M$0.037
15GPT-5.2 ProOpenAI74$21.0/M$168.0/M$1.277
16Kimi K3Moonshot72.18$2.9/M$14.3/M$0.118
17Gemini 3.7 FlashGoogle71.58$0.75/M$3.8/M$0.031
18Claude Opus 4.7Anthropic70.18$5.0/M$25.0/M$0.214
19GLM-5.3Z.ai68.77$1.4/M$4.4/M$0.042
20Gemini 3.8 FlashGoogle68.42$0.75/M$3.8/M$0.033
21GPT-5.2OpenAI67.4$1.8/M$14.0/M$0.117
22Claude Opus 4.6Anthropic65.96$5.0/M$25.0/M$0.227
23Grok 4.6SpaceXAI65.96$2.0/M$6.0/M$0.061
24Claude Sonnet 5Anthropic65.61$2.0/M$10.0/M$0.091
25Qwen3.7 MaxAlibaba64.56$2.5/M$7.5/M$0.077
26Gemini 3.5 FlashGoogle62.81$1.5/M$9.0/M$0.084
27Gemini 3.1 ProGoogle59.65$2.0/M$12.0/M$0.117
28GLM-5.2Z.ai59.21$1.4/M$4.4/M$0.049
29Gemini 3.6 FlashGoogle58.95$0.75/M$3.8/M$0.038
30DeepSeek-V4-FlashDeepSeek57.54$0.44/M$1.3/M$0.015
31Kimi K2.6Moonshot57.19$0.97/M$4.0/M$0.044
32Grok4.5SpaceXAI57.19$2.0/M$6.0/M$0.070
33GPT-5 ProOpenAI55.79$15.0/M$120.0/M$1.210
34GLM 5.3 FlashZ.ai55.79$0.15/M$0.50/M$0.006
35GPT-5+2 altsOpenAI55.44$1.3/M$10.0/M$0.101
36Kimi K2.7 CodeMoonshot54.04$1.9/M$8.0/M$0.092
37Gemini 3 Flash PreviewGoogle51.23$0.50/M$3.0/M$0.034
38GPT-5.4 mini+2 altsOpenAI51.23$0.75/M$4.5/M$0.051
39GPT-5 mini+2 altsOpenAI46.67$0.25/M$2.0/M$0.024
40Inkling-SmallThinking Machines46.32$0.30/M$1.2/M$0.016
41DeepSeek-V4-ProDeepSeek45.26$0.43/M$0.87/M$0.014
42GPT-5.4 nano+2 altsOpenAI44.91$0.20/M$1.3/M$0.016
43Grok 4.20SpaceXAI44.91$1.3/M$2.5/M$0.042
44Grok 4.3 BetaSpaceXAI42.81———
45Qwen3.6 PlusAlibaba38.25$0.50/M$3.0/M$0.046
46GLM-5.1Z.ai36.84$1.4/M$4.4/M$0.078
47O4 Mini+2 altsOpenAI36.14$1.1/M$4.4/M$0.076
48Qwen3.6 27BAlibaba35.09$0.60/M$3.6/M$0.060
49Qwen3.7 Plus PreviewAlibaba34.39$0.40/M$1.6/M$0.029
50Claude Opus 4.5Anthropic34.39$5.0/M$25.0/M$0.436
51InklingThinking Machines33.33$0.95/M$4.0/M$0.075
52O3+2 altsOpenAI33.33$2.0/M$8.0/M$0.150
53Qwen3.5 397B A17BAlibaba31.23$0.60/M$3.6/M$0.067
54Gemini 3.1 Flash Lite PreviewGoogle27.72$0.25/M$1.5/M$0.032
55GPT-5.5 InstantOpenAI26.32$5.0/M$30.0/M$0.665
56Gemini 3.5 Flash LiteGoogle25.96$0.30/M$2.5/M$0.054
57Gemini 2.5 ProGoogle24.56$1.3/M$10.0/M$0.229
58Claude Sonnet 4.5Anthropic23.86$3.0/M$15.0/M$0.377
59Qwen 3.6 FlashAlibaba22.46$0.25/M$1.5/M$0.039
60Qwen3.6 35B A3BAlibaba20.35$0.38/M$2.3/M$0.064
61GPT-5 nano+2 altsOpenAI20$0.05/M$0.40/M$0.011
62Qwen3.7 FlashAlibaba19.3$0.03/M$0.13/M$0.004
63Qwen3 Max InstructAlibaba18.95———
64O3 Mini+2 altsOpenAI18.6$1.1/M$4.4/M$0.148
65Qwen3.5 FlashAlibaba18.25$0.10/M$0.40/M$0.014
66O1+2 altsOpenAI14.74$15.0/M$60.0/M$2.544
67Claude Opus 4.1Anthropic12.63$15.0/M$75.0/M$3.563
68GPT-4.1 miniOpenAI6.67$0.40/M$1.6/M$0.150
69GPT-4.1OpenAI5.96$2.0/M$8.0/M$0.839
◆ 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: 69 models scored · 69 independently verified · 0 with source disagreement. How these tiers are assigned