Which model leads Claw-Eval Avg?

Q: Which model leads Claw-Eval Avg?

Claude Opus 4.5 tops the board at 76.6

Claw-Eval Avg Leaderboard — Model Scores

Pro Mode — raw scores, one benchmark under the lensofficial card harness

The field · 13 measured

48.576.6

13 models measured, most on the official card harness. Claude Opus 4.5 tops the board at 76.6.

13 measured·Lifecycle Updated Oct 9, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 13Bars show 48.5–76.6 · ◆ solid = first-party or better · outlined = arm's-length

01Claude Opus 4.5Anthropic76.63P$5.0/M$25.0/M$0.196

02Qwen3.6 27BAlibaba72.43P$0.60/M$3.6/M$0.029

03Qwen3.5 397B A17BAlibaba70.73P$0.60/M$3.6/M$0.030

04Qwen3.6 35B A3B+1 altAlibaba68.7$0.38/M$2.3/M$0.019

05Qwen3.5 9B+1 altAlibaba66.53$0.14/M$0.20/M$0.003

06Qwen3.5 35B A3BAlibaba65.4$0.25/M$2.0/M$0.017

07Qwen3.5 27B+1 altAlibaba64.3$0.30/M$2.4/M$0.021

08LFM2.5-2.6B+1 altLiquid AI62.85———

09Qwen3.5 4B+1 altAlibaba62.28$0.03/M$0.15/M$0.001

10Gemma 4 26B A4BGoogle58.8$0.07/M$0.34/M$0.003

11Gemma 4 E4B+1 altGoogle58.02$0.02/M$0.10/M$0.001

12Gemma 4 E2B+1 altGoogle53.14———

13Gemma 4 31B+1 altGoogle48.5$0.17/M$0.40/M$0.006

◆ 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: 13 models scored · 0 independently verified · 3 vendor cross-reference · 10 vendor-reported · 0 with source disagreement. How these tiers are assigned