Claude Sonnet 4.5 in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
B 1,420 ±5
Percentile 80 · #61
People prefer / Solves tasks
1,438 / 1,395
Human votes vs benchmarks, Elo-eq
Value
-2
$6 per 1M tokens, median of 34 providers · vendor $6
On your hardware
Closed weights: cloud only.
Score decomposition
B General 1,420 [1,415–1,424] Percentile 80 · #61 · components: 9
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| Epoch AI |
eci
|
147 ±0.83 | 1,413 ±6 | 34.6% | 2025-09-29 | — |
| LMArena |
text/overall
|
1,439 ±1.46 | 1,439 ±2 | 15.5% | 2026-10-02 | — |
| LMArena |
text/hard_prompts
|
1,462 ±1.84 | 1,451 ±2 | 15.5% | 2026-10-02 | — |
| LMArena |
text/instruction_following
|
1,459 ±2.35 | 1,469 ±2 | 7.7% | 2026-10-02 | — |
| LMArena |
text/expert
|
1,483 ±4.07 | 1,462 ±3 | 7.5% | 2026-10-02 | high-32k |
| LMArena |
text/longer_query
|
1,476 ±2.35 | 1,472 ±2 | 3.8% | 2026-10-02 | — |
| Epoch AI |
simpleqa_verified
decay ×0.61 |
0.31 ±0.01 | 1,394 ±5 | 3.8% | 2026-08-27 | 59k |
| LMArena |
text/multi_turn
|
1,465 ±2.87 | 1,458 ±3 | 3.8% | 2026-10-02 | — |
| Epoch AI |
gpqa_diamond
decay ×0.26 |
0.82 ±0.03 | 1,423 ±13 | 1.1% | 2025-10-28 | 59k |
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
B Coding 1,420 [1,416–1,423] Percentile 77 · #64 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
webdev/overall
|
1,393 ±3.92 | 1,419 ±1 | 39.6% | 2026-10-01 | high-32k |
| LMArena |
text/coding
|
1,488 ±2.48 | 1,460 ±2 | 39.0% | 2026-10-02 | high-32k |
| Epoch AI |
swe_bench_verified
decay ×0.37 |
0.71 ±0.02 | 1,439 ±11 | 13.1% | 2026-02-05 | — |
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 Agents 1,404 [1,393–1,416] Percentile 23 · #45 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| τ²-bench (Sierra) |
tau2/banking_knowledge
decay ×0.76 |
0.25 ±0.04 | 1,456 ±7 | 79.6% | 2026-02-24 | enabled |
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
B Math 1,398 [1,393–1,404] Percentile 69 · #80 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,448 ±4.31 | 1,439 ±4 | 50.8% | 2026-10-02 | high-32k |
| Epoch AI |
frontiermath_t1_3
decay ×0.86 |
0.24 ±0.03 | 1,373 ±6 | 27.7% | 2026-06-11 | 32k |
| Epoch AI |
frontiermath_t4
decay ×0.86 |
0.02 ±0.02 | 1,396 ±4 | 14.7% | 2026-06-11 | 32k |
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
A Writing 1,455 [1,451–1,459] Percentile 89 · #31 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,442 ±3.07 | 1,471 ±3 | 62.2% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,476 ±2.35 | 1,472 ±2 | 31.8% | 2026-10-02 | — |
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
B Language: 繁體中文 1,408 [1,402–1,414] Percentile 62 · #85 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,459 ±4.13 | 1,421 ±3 | 93.8% | 2026-10-02 | high-32k |
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
B Language: Deutsch 1,410 [1,397–1,423] Percentile 64 · #52 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/german
|
1,428 ±8.50 | 1,426 ±7 | 92.8% | 2026-10-02 | high-32k |
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
B Language: English 1,427 [1,423–1,431] Percentile 79 · #58 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,450 ±2.02 | 1,441 ±2 | 94.0% | 2026-10-02 | — |
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
B Language: Español 1,434 [1,423–1,445] Percentile 79 · #32 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,456 ±6.37 | 1,451 ±6 | 93.3% | 2026-10-02 | — |
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
B Language: Français 1,423 [1,411–1,436] Percentile 68 · #44 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,458 ±7.46 | 1,440 ±7 | 93.0% | 2026-10-02 | — |
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
B Language: 日本語 1,403 [1,388–1,418] Percentile 55 · #58 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/japanese
|
1,390 ±11 | 1,419 ±8 | 92.4% | 2026-10-02 | — |
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
B Language: 한국어 1,421 [1,409–1,433] Percentile 70 · #41 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/korean
|
1,403 ±8.70 | 1,438 ±7 | 93.1% | 2026-10-02 | — |
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
B Language: Polski 1,412 [1,402–1,422] Percentile 54 · #61 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/polish
|
1,423 ±6.46 | 1,426 ±6 | 93.4% | 2026-10-02 | high-32k |
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
B Language: Русский 1,425 [1,420–1,431] Percentile 78 · #52 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,436 ±3.30 | 1,440 ±3 | 93.9% | 2026-10-02 | — |
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
The same in a table
| Model | 2026-10 |
|---|---|
| Claude Sonnet 4.5 | 1,420 B |
Badge
Show the model's fedi-index on your site or in a README. The badge updates itself; it links to this page.
Data CC BY 4.0: keep the link to fedi.software.