gpt-oss-20b in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
C 1,293 ±9
Percentile 34 · #195
People prefer / Solves tasks
1,273 / 1,325
Human votes vs benchmarks, Elo-eq
Value
-26
$0.0875 per 1M tokens, median of 30 providers
On your hardware
- GPU 12 GB #13 · MXFP4 · ≈53 tok/s
- GPU 24 GB #19 · MXFP4 · ≈180 tok/s
- Mac 64 GB #17 · MXFP4 · ≈80 tok/s
- 128 GB unified memory #32 · MXFP4 · ≈59 tok/s
- CPU, 32 GB RAM #13 · MXFP4 · ≈7 tok/s
Score decomposition
C General 1,293 [1,284–1,301] Percentile 34 · #195 · components: 8
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| Epoch AI |
eci
|
138 ±1.71 | 1,342 ±14 | 31.7% | 2025-08-05 | — |
| LMArena |
text/overall
|
1,288 ±3.27 | 1,288 ±3 | 18.2% | 2026-10-02 | — |
| LMArena |
text/hard_prompts
|
1,274 ±4.57 | 1,281 ±4 | 17.8% | 2026-10-02 | — |
| LMArena |
text/instruction_following
|
1,236 ±6.25 | 1,258 ±6 | 8.3% | 2026-10-02 | — |
| LMArena |
text/expert
|
1,257 ±14 | 1,278 ±11 | 6.1% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,251 ±6.78 | 1,261 ±6 | 4.0% | 2026-10-02 | — |
| LMArena |
text/multi_turn
|
1,268 ±7.39 | 1,279 ±7 | 4.0% | 2026-10-02 | — |
| Epoch AI |
gpqa_diamond
decay ×0.26 |
0.61 ±0.02 | 1,323 ±12 | 1.6% | 2026-08-27 | 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
C Coding 1,288 [1,278–1,298] Percentile 30 · #194 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/coding
|
1,307 ±6.71 | 1,298 ±6 | 89.2% | 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 Math 1,296 [1,278–1,314] Percentile 35 · #167 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,317 ±11 | 1,313 ±11 | 83.9% | 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
C Writing 1,236 [1,224–1,247] Percentile 27 · #195 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,201 ±9.11 | 1,225 ±9 | 58.2% | 2026-10-02 | — |
| LMArena |
text/longer_query
|
1,251 ±6.78 | 1,261 ±6 | 34.2% | 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: 繁體中文 1,295 [1,279–1,312] Percentile 27 · #163 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,311 ±12 | 1,304 ±9 | 92.0% | 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: English 1,280 [1,271–1,289] Percentile 33 · #184 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,303 ±4.63 | 1,286 ±5 | 93.5% | 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
D Language: Polski 1,275 [1,258–1,292] Percentile 2 · #129 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/polish
|
1,251 ±11 | 1,281 ±10 | 92.0% | 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: Русский 1,290 [1,273–1,308] Percentile 27 · #171 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,279 ±11 | 1,298 ±10 | 91.9% | 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 |
|---|---|
| gpt-oss-20b | 1,293 C |
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.