Skip to content
fedi.software

Llama 2 7B Chat in fedi-index

The score in every slice, what it is made of, value, local run and history.

fedi-index · General

D 1,057 ±9

Percentile 6 · #278

People prefer / Solves tasks

1,062 / 1,030

Human votes vs benchmarks, Elo-eq

Value

-253

$0.125 per 1M tokens, median of 2 providers

On your hardware

Closed weights: cloud only.

Score decomposition

D General 1,057 [1,048–1,066] Percentile 6 · #278 · components: 7
Source Component Value Elo-eq Weight Measured Variant
LMArena text/overall 1,053 ±3.55 1,053 ±4 25.2% 2026-11-11 —
LMArena text/hard_prompts 1,009 ±5.50 1,042 ±5 24.0% 2026-11-11 —
LMArena text/instruction_following 1,006 ±4.99 1,039 ±5 12.1% 2026-11-11 —
Epoch AI eci 83.38 ±6.00 912 ±47 8.5% 2023-07-18 —
LMArena text/expert 1,036 ±14 1,097 ±12 8.3% 2026-11-11 —
LMArena text/multi_turn 1,029 ±7.67 1,062 ±7 5.5% 2026-11-11 —
LMArena text/longer_query 999 ±10 1,026 ±9 4.8% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Coding 1,049 [1,038–1,060] Percentile 4 · #275 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/coding 1,002 ±7.13 1,026 ±6 87.2% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Math 1,070 [1,058–1,081] Percentile 7 · #239 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/math 1,042 ±7.08 1,050 ±7 86.9% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Writing 1,057 [1,046–1,067] Percentile 7 · #245 · components: 2 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/creative_writing 1,033 ±6.92 1,054 ±7 64.4% 2026-11-11 —
LMArena text/longer_query 999 ±10 1,026 ±9 28.2% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Language: 繁體中文 1,047 [1,033–1,062] Percentile 2 · #215 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/chinese 973 ±10 1,035 ±8 92.6% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Language: Deutsch 1,056 [1,030–1,083] Percentile 1 · #144 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/german 977 ±18 1,038 ±15 88.6% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Language: English 1,075 [1,068–1,083] Percentile 10 · #242 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/english 1,096 ±4.03 1,067 ±4 93.7% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

D Language: Русский 1,056 [1,037–1,075] Percentile 2 · #225 · components: 1 provisional
Source Component Value Elo-eq Weight Measured Variant
LMArena text/russian 995 ±12 1,043 ±11 91.4% 2026-11-11 —

Weight is the share of the component in the score (the rest is the prior that pulls models with little data towards the population median). Methodology

History

fedi-index by month
1,055 1,060 1,065 1,070 2026-10 2026-11 Llama 2 7B Chat · 2026-10 · 1,066Llama 2 7B Chat · 2026-11 · 1,057 Llama 2 7B Chat
The same in a table
Model 2026-102026-11
Llama 2 7B Chat 1,0661,057 D

Badge

Show the model's fedi-index on your site or in a README. The badge updates itself; it links to this page.

fedi-index: Llama 2 7B Chat

Data CC BY 4.0: keep the link to fedi.software.