o3 in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
B 1,392 ±5
Percentile 64 · #107
People prefer / Solves tasks
1,386 / 1,400
Human votes vs benchmarks, Elo-eq
Value
-17
$3.50 per 1M tokens, median of 18 providers · vendor $3.50
On your hardware
Closed weights: cloud only.
Score decomposition
B General 1,392 [1,386–1,396] Percentile 64 · #107 · components: 9
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| Epoch AI |
eci
|
147 ±0.92 | 1,413 ±7 | 33.9% | 2025-04-16 | — |
| LMArena |
text/overall
|
1,410 ±1.84 | 1,410 ±2 | 15.7% | 2026-10-02 | — |
| LMArena |
text/hard_prompts
|
1,402 ±2.43 | 1,397 ±2 | 15.6% | 2026-10-02 | — |
| LMArena |
text/instruction_following
|
1,368 ±2.83 | 1,382 ±3 | 7.7% | 2026-10-02 | — |
| LMArena |
text/expert
|
1,403 ±5.77 | 1,397 ±5 | 7.3% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,371 ±3.13 | 1,374 ±3 | 3.8% | 2026-10-02 | — |
| Epoch AI |
simpleqa_verified
decay ×0.61 |
0.49 ±0.02 | 1,452 ±5 | 3.8% | 2026-08-27 | high |
| LMArena |
text/multi_turn
|
1,405 ±3.45 | 1,403 ±3 | 3.8% | 2026-10-02 | — |
| Epoch AI |
gpqa_diamond
decay ×0.26 |
0.82 ±0.02 | 1,420 ±10 | 1.3% | 2025-04-16 | high |
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 1,373 [1,368–1,378] Percentile 53 · #130 · components: 2
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/coding
|
1,408 ±3.16 | 1,388 ±3 | 78.7% | 2026-10-02 | — |
| Epoch AI |
swe_bench_verified
decay ×0.37 |
0.62 ±0.02 | 1,393 ±11 | 12.8% | 2026-02-12 | medium |
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 1,392 [1,386–1,400] Percentile 65 · #90 · components: 2
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,426 ±5.10 | 1,418 ±5 | 51.0% | 2026-10-02 | — |
| Epoch AI |
frontiermath_t1_3
decay ×0.86 |
0.33 ±0.03 | 1,393 ±6 | 41.9% | 2026-08-27 | high |
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
C Vision 1,384 [1,379–1,388] Percentile 42 · #66 · components: 3 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
vision/overall
|
1,214 ±3.34 | 1,397 ±3 | 63.7% | 2026-10-02 | — |
| LMArena |
vision/ocr
|
1,218 ±4.19 | 1,393 ±4 | 15.4% | 2026-10-02 | — |
| LMArena |
vision/diagram
|
1,219 ±6.13 | 1,392 ±5 | 14.6% | 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 Writing 1,371 [1,366–1,376] Percentile 55 · #122 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,358 ±3.72 | 1,386 ±4 | 62.0% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,371 ±3.13 | 1,374 ±3 | 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 1,392 [1,384–1,399] Percentile 53 · #106 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,438 ±5.10 | 1,404 ±4 | 93.7% | 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: Deutsch 1,404 [1,391–1,417] Percentile 61 · #57 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/german
|
1,420 ±8.18 | 1,419 ±7 | 92.9% | 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
B Language: English 1,393 [1,388–1,398] Percentile 60 · #109 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,416 ±2.40 | 1,405 ±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
C Language: Español 1,380 [1,364–1,395] Percentile 33 · #100 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,395 ±9.17 | 1,394 ±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
C Language: Français 1,397 [1,378–1,415] Percentile 44 · #76 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,430 ±11 | 1,414 ±10 | 91.6% | 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 1,414 [1,401–1,426] Percentile 65 · #46 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/japanese
|
1,403 ±8.94 | 1,429 ±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: 한국어 1,398 [1,386–1,411] Percentile 52 · #66 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/korean
|
1,371 ±9.34 | 1,413 ±7 | 92.9% | 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
B Language: Polski 1,424 [1,416–1,433] Percentile 68 · #43 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/polish
|
1,438 ±5.48 | 1,439 ±5 | 93.6% | 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: Русский 1,400 [1,392–1,407] Percentile 61 · #92 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,407 ±4.48 | 1,413 ±4 | 93.7% | 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 |
|---|---|
| o3 | 1,392 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.