Electricity Bench

Compare models

Every graded pair, side by side: per-suite grades, capability, task time, approx. task cost, and each agent's plan cost. Weighing the plans instead? Compare subscriptions.

Choose two agents

Select each agent by name. Native select controls support keyboard typeahead.

Browse the matrix

Each cell reads row grade / column grade and links to that pair's full head-to-head page. Rows and columns are in leaderboard order, and the board scrolls sideways when it is wider than the screen.

Claude Opus 5.5Claude Fable 5.1GPT-6.1 SolGrok 4.7GPT-6 AstraClaude Sonnet 5.5Gemini 3.8 FlashKimi K3 256KGPT-6 LunaGPT-5.6 TerraComposer 2.5Gemini 3.1 ProClaude Haiku 4.5
Claude Opus 5.5B-B-/B-B-/CB-/CB-/CB-/C-B-/C-B-/D+B-/DB-/DB-/DB-/FB-/F
Claude Fable 5.1B-B-/CB-/CB-/CB-/C-B-/C-B-/D+B-/DB-/DB-/DB-/FB-/F
GPT-6.1 SolCC/CC/CC/C-C/C-C/D+C/DC/DC/DC/FC/F
Grok 4.7CC/CC/C-C/C-C/D+C/DC/DC/DC/FC/F
GPT-6 AstraCC/C-C/C-C/D+C/DC/DC/DC/FC/F
Claude Sonnet 5.5C-C-/C-C-/D+C-/DC-/DC-/DC-/FC-/F
Gemini 3.8 FlashC-C-/D+C-/DC-/DC-/DC-/FC-/F
Kimi K3 256KD+D+/DD+/DD+/DD+/FD+/F
GPT-6 LunaDD/DD/DD/FD/F
GPT-5.6 TerraDD/DD/FD/F
Composer 2.5DD/FD/F
Gemini 3.1 ProFF/F
Claude Haiku 4.5F
Show previous versions (10)

These agents ran a model version that has since been replaced. Their last published overall grade still stands, and every comparison page they appear in still works - but they are no longer graded week to week, so their figures age.

Claude Opus 5.5Claude Fable 5.1GPT-6.1 SolGrok 4.7GPT-6 AstraClaude Sonnet 5.5Gemini 3.8 FlashKimi K3 256KGPT-6 LunaGPT-5.6 TerraComposer 2.5Gemini 3.1 ProClaude Haiku 4.5Claude Fable 5Claude Opus 5Claude Sonnet 5Gemini 3.6 FlashGemini 3.7 FlashGPT-5.6 LunaGPT-5.6 SolGPT-6 SolGrok 4.5Grok 4.6
Claude Fable 5CClaude Fable 5 was superseded by Claude Fable 5.1.C/B-C/B-C/CC/CC/CC/C-C/C-C/D+C/DC/DC/DC/FC/FC/CC/DC/CC/D+C/DC/C-C/C-C/C-C/C-
Claude Opus 5CClaude Opus 5 was superseded by Claude Opus 5.5.C/B-C/B-C/CC/CC/CC/C-C/C-C/D+C/DC/DC/DC/FC/FC/CC/DC/CC/D+C/DC/C-C/C-C/C-C/C-
Claude Sonnet 5DClaude Sonnet 5 is no longer on the leaderboard.D/B-D/B-D/CD/CD/CD/C-D/C-D/D+D/DD/DD/DD/FD/FD/CD/CD/CD/D+D/DD/C-D/C-D/C-D/C-
Gemini 3.6 FlashCGemini 3.6 Flash was superseded by Gemini 3.8 Flash.C/B-C/B-C/CC/CC/CC/C-C/C-C/D+C/DC/DC/DC/FC/FC/CC/CC/DC/D+C/DC/C-C/C-C/C-C/C-
Gemini 3.7 FlashD+Gemini 3.7 Flash was superseded by Gemini 3.8 Flash.D+/B-D+/B-D+/CD+/CD+/CD+/C-D+/C-D+/D+D+/DD+/DD+/DD+/FD+/FD+/CD+/CD+/DD+/CD+/DD+/C-D+/C-D+/C-D+/C-
GPT-5.6 LunaDGPT-5.6 Luna was superseded by GPT-6 Luna.D/B-D/B-D/CD/CD/CD/C-D/C-D/D+D/DD/DD/DD/FD/FD/CD/CD/DD/CD/D+D/C-D/C-D/C-D/C-
GPT-5.6 SolC-GPT-5.6 Sol was superseded by GPT-6.1 Sol.C-/B-C-/B-C-/CC-/CC-/CC-/C-C-/C-C-/D+C-/DC-/DC-/DC-/FC-/FC-/CC-/CC-/DC-/CC-/D+C-/DC-/C-C-/C-C-/C-
GPT-6 SolC-GPT-6 Sol was superseded by GPT-6.1 Sol.C-/B-C-/B-C-/CC-/CC-/CC-/C-C-/C-C-/D+C-/DC-/DC-/DC-/FC-/FC-/CC-/CC-/DC-/CC-/D+C-/DC-/C-C-/C-C-/C-
Grok 4.5C-Grok 4.5 was superseded by Grok 4.7.C-/B-C-/B-C-/CC-/CC-/CC-/C-C-/C-C-/D+C-/DC-/DC-/DC-/FC-/FC-/CC-/CC-/DC-/CC-/D+C-/DC-/C-C-/C-C-/C-
Grok 4.6C-Grok 4.6 was superseded by Grok 4.7.C-/B-C-/B-C-/CC-/CC-/CC-/C-C-/C-C-/D+C-/DC-/DC-/DC-/FC-/FC-/CC-/CC-/DC-/CC-/D+C-/DC-/C-C-/C-C-/C-
Browse all 78 pair cards

Claude Fable 5.1 vs Claude Haiku 4.5

B-vsF

Both on claude-code

Claude Fable 5.1 vs Claude Opus 5.5

B-vsB-

Both on claude-code

Claude Fable 5.1 vs Claude Sonnet 5.5

B-vsC-

Both on claude-code

Claude Fable 5.1 vs Composer 2.5

B-vsD

claude-code vs cursor-cli

Claude Fable 5.1 vs GPT-5.6 Terra

B-vsD

claude-code vs codex-cli

Claude Fable 5.1 vs GPT-6 Astra

B-vsC

claude-code vs codex-cli

Claude Fable 5.1 vs GPT-6 Luna

B-vsD

claude-code vs codex-cli

Claude Fable 5.1 vs GPT-6.1 Sol

B-vsC

claude-code vs codex-cli

Claude Fable 5.1 vs Grok 4.7

B-vsC

claude-code vs grok-cli

Claude Fable 5.1 vs Kimi K3 256K

B-vsD+

claude-code vs kimi-cli

Claude Haiku 4.5 vs Claude Opus 5.5

FvsB-

Both on claude-code

Claude Opus 5.5 vs Claude Sonnet 5.5

B-vsC-

Both on claude-code

Claude Opus 5.5 vs Composer 2.5

B-vsD

claude-code vs cursor-cli

Claude Opus 5.5 vs GPT-5.6 Terra

B-vsD

claude-code vs codex-cli

Claude Opus 5.5 vs GPT-6 Astra

B-vsC

claude-code vs codex-cli

Claude Opus 5.5 vs GPT-6 Luna

B-vsD

claude-code vs codex-cli

Claude Opus 5.5 vs GPT-6.1 Sol

B-vsC

claude-code vs codex-cli

Claude Opus 5.5 vs Grok 4.7

B-vsC

claude-code vs grok-cli

Claude Opus 5.5 vs Kimi K3 256K

B-vsD+

claude-code vs kimi-cli

Gemini 3.1 Pro vs Claude Fable 5.1

FvsB-

antigravity-cli vs claude-code

Gemini 3.1 Pro vs Claude Opus 5.5

FvsB-

antigravity-cli vs claude-code

Gemini 3.8 Flash vs Claude Fable 5.1

C-vsB-

antigravity-cli vs claude-code

Gemini 3.8 Flash vs Claude Opus 5.5

C-vsB-

antigravity-cli vs claude-code

Claude Haiku 4.5 vs GPT-6 Astra

FvsC

claude-code vs codex-cli

Claude Haiku 4.5 vs GPT-6.1 Sol

FvsC

claude-code vs codex-cli

Claude Haiku 4.5 vs Grok 4.7

FvsC

claude-code vs grok-cli

Claude Sonnet 5.5 vs GPT-6 Astra

C-vsC

claude-code vs codex-cli

Claude Sonnet 5.5 vs GPT-6.1 Sol

C-vsC

claude-code vs codex-cli

Claude Sonnet 5.5 vs Grok 4.7

C-vsC

claude-code vs grok-cli

Composer 2.5 vs Grok 4.7

DvsC

cursor-cli vs grok-cli

Gemini 3.1 Pro vs GPT-6 Astra

FvsC

antigravity-cli vs codex-cli

Gemini 3.1 Pro vs GPT-6.1 Sol

FvsC

antigravity-cli vs codex-cli

Gemini 3.1 Pro vs Grok 4.7

FvsC

antigravity-cli vs grok-cli

Gemini 3.8 Flash vs GPT-6 Astra

C-vsC

antigravity-cli vs codex-cli

Gemini 3.8 Flash vs GPT-6.1 Sol

C-vsC

antigravity-cli vs codex-cli

Gemini 3.8 Flash vs Grok 4.7

C-vsC

antigravity-cli vs grok-cli

GPT-5.6 Terra vs GPT-6 Astra

DvsC

Both on codex-cli

GPT-5.6 Terra vs GPT-6.1 Sol

DvsC

Both on codex-cli

GPT-5.6 Terra vs Grok 4.7

DvsC

codex-cli vs grok-cli

GPT-6 Astra vs Composer 2.5

CvsD

codex-cli vs cursor-cli

GPT-6 Astra vs GPT-6 Luna

CvsD

Both on codex-cli

GPT-6 Astra vs Grok 4.7

CvsC

codex-cli vs grok-cli

GPT-6 Astra vs Kimi K3 256K

CvsD+

codex-cli vs kimi-cli

GPT-6 Luna vs Grok 4.7

DvsC

codex-cli vs grok-cli

GPT-6.1 Sol vs Composer 2.5

CvsD

codex-cli vs cursor-cli

GPT-6.1 Sol vs GPT-6 Astra

CvsC

Both on codex-cli

GPT-6.1 Sol vs GPT-6 Luna

CvsD

Both on codex-cli

GPT-6.1 Sol vs Grok 4.7

CvsC

codex-cli vs grok-cli

GPT-6.1 Sol vs Kimi K3 256K

CvsD+

codex-cli vs kimi-cli

Grok 4.7 vs Kimi K3 256K

CvsD+

grok-cli vs kimi-cli

Claude Haiku 4.5 vs Claude Sonnet 5.5

FvsC-

Both on claude-code

Claude Sonnet 5.5 vs Composer 2.5

C-vsD

claude-code vs cursor-cli

Claude Sonnet 5.5 vs GPT-5.6 Terra

C-vsD

claude-code vs codex-cli

Claude Sonnet 5.5 vs GPT-6 Luna

C-vsD

claude-code vs codex-cli

Claude Sonnet 5.5 vs Kimi K3 256K

C-vsD+

claude-code vs kimi-cli

Gemini 3.1 Pro vs Claude Sonnet 5.5

FvsC-

antigravity-cli vs claude-code

Gemini 3.1 Pro vs Gemini 3.8 Flash

FvsC-

Both on antigravity-cli

Gemini 3.8 Flash vs Claude Haiku 4.5

C-vsF

antigravity-cli vs claude-code

Gemini 3.8 Flash vs Claude Sonnet 5.5

C-vsC-

antigravity-cli vs claude-code

Gemini 3.8 Flash vs Composer 2.5

C-vsD

antigravity-cli vs cursor-cli

Gemini 3.8 Flash vs GPT-5.6 Terra

C-vsD

antigravity-cli vs codex-cli

Gemini 3.8 Flash vs GPT-6 Luna

C-vsD

antigravity-cli vs codex-cli

Gemini 3.8 Flash vs Kimi K3 256K

C-vsD+

antigravity-cli vs kimi-cli

Claude Haiku 4.5 vs Kimi K3 256K

FvsD+

claude-code vs kimi-cli

Composer 2.5 vs Kimi K3 256K

DvsD+

cursor-cli vs kimi-cli

Gemini 3.1 Pro vs Kimi K3 256K

FvsD+

antigravity-cli vs kimi-cli

GPT-5.6 Terra vs Kimi K3 256K

DvsD+

codex-cli vs kimi-cli

GPT-6 Luna vs Kimi K3 256K

DvsD+

codex-cli vs kimi-cli

Claude Haiku 4.5 vs Composer 2.5

FvsD

claude-code vs cursor-cli

Claude Haiku 4.5 vs GPT-5.6 Terra

FvsD

claude-code vs codex-cli

Claude Haiku 4.5 vs GPT-6 Luna

FvsD

claude-code vs codex-cli

Gemini 3.1 Pro vs Composer 2.5

FvsD

antigravity-cli vs cursor-cli

Gemini 3.1 Pro vs GPT-5.6 Terra

FvsD

antigravity-cli vs codex-cli

Gemini 3.1 Pro vs GPT-6 Luna

FvsD

antigravity-cli vs codex-cli

GPT-5.6 Terra vs Composer 2.5

DvsD

codex-cli vs cursor-cli

GPT-5.6 Terra vs GPT-6 Luna

DvsD

Both on codex-cli

GPT-6 Luna vs Composer 2.5

DvsD

codex-cli vs cursor-cli

Gemini 3.1 Pro vs Claude Haiku 4.5

FvsF

antigravity-cli vs claude-code