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 · 70 measured
793.9

70 models measured, most on the official card harness. Claude Opus 5.5 tops the board at 93.9.

70 measured·3 new this week·Lifecycle Updated Oct 8, 2026Score basis
#◆Vendor$/M in$/M out
Leader — ignites whiteField — fill cools with scoreBars show 7–93.9 · ◆ solid = first-party or better · outlined = arm's-length
01Claude Opus 5.5+2 altsAnthropic93.9$4.0/M$20.0/M$0.128
02Sonnet 5.5+2 altsAnthropic90.3$2.0/M$10.0/M$0.066
03Claude Opus 5+2 altsAnthropic89.5$5.0/M$25.0/M$0.168
04Claude Fable 5.1+1 altAnthropic89.1$10.0/M$50.0/M$0.337
05Claude Mythos 5.1Anthropic89.1$10.0/M$50.0/M$0.337
06Claude Mythos PreviewAnthropic87.3———
07Claude Mythos 5+1 altAnthropic86.6$10.0/M$50.0/M$0.346
08Claude Opus 4.8+2 altsAnthropic84.4$5.0/M$25.0/M$0.178
09Haiku 5.5NEW+1 altAnthropic83.7$0.10/M$0.50/M$0.004
10Hy4Tencent82.9$0.83/M$2.5/M$0.020
11Qwen3.8 Flash Next+1 altAlibaba81$0.15/M$0.47/M$0.004
12Claude Opus 4.7+2 altsAnthropic80.5$5.0/M$25.0/M$0.186
13Laguna S 2.1poolside78.5$0.09/M$0.18/M$0.002
14Qwen3.7 MaxAlibaba78.3$2.5/M$7.5/M$0.064
15Claude Sonnet 5+3 altsAnthropic78.3$2.0/M$10.0/M$0.077
16Claude Opus 4.6+8 altsAnthropic77.83$5.0/M$25.0/M$0.193
17Gemini 3.1 Pro+1 altGoogle76.93P$2.0/M$12.0/M$0.091
18Kimi K2.6+8 altsMoonshot76.7$0.95/M$4.0/M$0.032
19MiniMax M2.7+1 altMiniMax76.5$0.21/M$0.84/M$0.007
20DeepSeek-V4-Pro+8 altsDeepSeek76.2$0.43/M$0.87/M$0.009
21Qwen3.7 Plus Preview+1 altAlibaba75.8$0.40/M$1.6/M$0.013
22Hy3-previewTencent75.8$0.14/M$0.56/M$0.005
23Qwen3.8 27BAlibaba73.8$0.50/M$3.0/M$0.024
24Qwen3.6 PlusAlibaba73.8$0.50/M$3.0/M$0.024
25GLM-5.1+4 altsZ.ai73.3$1.4/M$4.4/M$0.039
26DeepSeek-V4-Flash+1 altDeepSeek73.3$0.44/M$1.3/M$0.012
27Kimi K2.5+2 altsMoonshot73$0.45/M$2.3/M$0.018
28MiniMax M2.1+1 altMiniMax72.5$0.30/M$1.2/M$0.010
29Ling-3.0-flashInclusionAI72.4$0.07/M$0.22/M$0.002
30GPT-5.4+1 altOpenAI71.73P$2.5/M$15.0/M$0.122
31MiMo V2 ProXiaomi71.7———
32MiMo V2 FlashXiaomi71.7$0.10/M$0.30/M$0.003
33Qwen3.6 27B+1 altAlibaba71.3$0.60/M$3.6/M$0.029
34DeepSeek-V3.2+6 altsDeepSeek70.2$0.28/M$0.42/M$0.005
35Qwen3.5 27B+1 altAlibaba69.3$0.30/M$2.4/M$0.019
36Qwen3.5 397B A17B+2 altsAlibaba69.3$0.60/M$3.6/M$0.030
37Nemotron 3 Ultra 550B A55B+2 altsNVIDIA67.7$0.60/M$2.5/M$0.023
38Claude Haiku 4.5NEWAnthropic67.4$1.0/M$5.0/M$0.045
39Qwen3.6 35B A3B+3 altsAlibaba67.2$0.38/M$2.3/M$0.020
40Claude Sonnet 4.5+7 altsAnthropic673P$3.0/M$15.0/M$0.134
41GPT-5.2+2 altsOpenAI66.73P$1.8/M$14.0/M$0.118
42GLM-4.7+1 altZ.ai66.7$0.60/M$2.2/M$0.021
43Qwen3 Max (Reasoning)Alibaba66.7$1.2/M$6.0/M$0.054
44MAI-Code-1-FlashMicrosoft65.5———
45Gemini 3 ProGoogle653P$2.0/M$12.0/M$0.108
46Laguna-XS-2.1poolside63.1$0.06/M$0.12/M$0.001
47Devstral 2Mistral61.33P$0.40/M$2.0/M$0.020
48Kimi K2 Thinking+1 altMoonshot61.13P$0.60/M$2.5/M$0.025
49Qwen3.5 35B A3BAlibaba60.3$0.25/M$2.0/M$0.019
50DeepSeek-V3.2-ExpDeepSeek57.9$0.28/M$0.42/M$0.006
51Claude Sonnet 4Anthropic56.93P$3.0/M$15.0/M$0.158
52MiniMax M2+3 altsMiniMax56.5$0.30/M$1.2/M$0.013
53Devstral Small 2Mistral55.73P———
54GPT-5OpenAI55.33P$1.3/M$10.0/M$0.102
55Qwen3 Coder 480B A35B InstructAlibaba54.7$1.5/M$7.5/M$0.082
56Qwen3 Coder PlusAlibaba54.73P$1.0/M$5.0/M$0.055
57DeepSeek-V3.1DeepSeek54.5$0.56/M$1.7/M$0.021
58GLM-4.6+1 altZ.ai53.83P$0.57/M$2.2/M$0.026
59Gemma 4 31B+1 altGoogle51.7$0.17/M$0.40/M$0.005
60Nemotron 3 SuperNVIDIA49.83P———
61Kimi K2 Instruct+2 altsMoonshot47.3$0.57/M$2.3/M$0.030
62Nemotron 3 Super 120B A12BNVIDIA45.78$0.30/M$0.90/M$0.013
63GPT Oss 20bOpenAI41.933P$0.07/M$0.18/M$0.003
64Granite 4.2 30B+1 altIBM41.89$0.16/M$0.65/M$0.010
65Nemotron 3.5 Lightning+1 altNVIDIA39.33$0.06/M$0.20/M$0.003
66LongCat Flash LiteMeituan38.1———
67Granite 4.2 8B+2 altsIBM30.78$0.06/M$0.25/M$0.005
68Gemma 4 26B A4B+1 altGoogle17.3$0.07/M$0.34/M$0.012
69Nemotron 3 NanoNVIDIA14.073P———
70Claude Opus 4.5NEW+6 altsAnthropic7$5.0/M$25.0/M$2.143
◆ 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: 70 models scored · 2 independently verified · 11 vendor cross-reference · 57 vendor-reported · 4 with source disagreement. How these tiers are assigned