Which model leads AlpacaEval-2.0?

Q: Which model leads AlpacaEval-2.0?

Granite 3.3 8B Instruct tops the board at 62.68

AlpacaEval-2.0 Leaderboard — 16 Models Ranked

Pro Mode — raw scores, one benchmark under the lensthird-party eval harness

The field · 16 measured

15.3562.68

16 models measured, most on the third-party eval harness. Granite 3.3 8B Instruct tops the board at 62.68.

16 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 16Bars show 15.35–62.68 · ◆ solid = first-party or better · outlined = arm's-length

01Granite 3.3 8B Instruct+1 altIBM62.68VENDOR$0.03/M$0.25/M$0.002

02Granite 3.2 8B InstructIBM61.19———

03Granite 4.1 30BIBM56.16———

04Hunyuan Large InstructTencent51.8VENDOR———

05DeepSeek-V2.5+1 altDeepSeek50.5VENDOR———

06Granite 4.1 8BIBM50.08$0.05/M$0.10/M$0.001

07Granite 3.3 2B InstructIBM43.45———

08Llama 3.1 Instruct 405BMeta39.3VENDOR———

09Granite 4.1 3BIBM38.57———

10Granite 3.2 2B InstructIBM34.51———

11Llama 3.1 70B InstructMeta34.3VENDOR$0.56/M$0.56/M$0.016

12Granite 3.1 8B InstructIBM30.34———

13Llama 3.1 8B InstructMeta27.22$0.02/M$0.04/M$0.001

14Granite 3.1 2B InstructIBM27.17———

15DeepSeek-R1-Distill-Llama-8BDeepSeek21.85$0.05/M$0.05/M$0.002

16DeepSeek-R1-Distill-Qwen-7BDeepSeek15.35$0.15/M$0.15/M$0.010

◆ 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: 16 models scored · 0 independently verified · 10 vendor cross-reference · 6 vendor-reported · 0 with source disagreement. How these tiers are assigned