GLM 5.2 in fedi-index
The score in every slice, what it is made of, value, local run and history.
fedi-index · General
A 1,442 ±5
Percentile 88 · #37
People prefer / Solves tasks
1,453 / 1,427
Human votes vs benchmarks, Elo-eq
Value
+45
$2.15 per 1M tokens, median of 73 providers · vendor $2.15
On your hardware
Does not fit the reference setups.
Score decomposition
A General 1,442 [1,437–1,447] Percentile 88 · #37 · components: 9
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| Epoch AI |
eci
|
152 ±1.02 | 1,452 ±8 | 32.9% | 2026-06-16 | — |
| LMArena |
text/overall
|
1,470 ±2.18 | 1,470 ±2 | 15.8% | 2026-10-02 | max |
| LMArena |
text/hard_prompts
|
1,480 ±2.49 | 1,467 ±2 | 15.8% | 2026-10-02 | max |
| LMArena |
text/instruction_following
|
1,465 ±2.97 | 1,475 ±3 | 7.8% | 2026-10-02 | max |
| LMArena |
text/expert
|
1,486 ±4.61 | 1,465 ±4 | 7.6% | 2026-10-02 | max |
| LMArena |
text/longer_query
|
1,480 ±2.82 | 1,475 ±3 | 3.9% | 2026-10-02 | max |
| Epoch AI |
simpleqa_verified
decay ×0.61 |
0.34 ±0.02 | 1,405 ±5 | 3.9% | 2026-08-27 | max |
| LMArena |
text/multi_turn
|
1,468 ±3.91 | 1,461 ±4 | 3.8% | 2026-10-02 | max |
| Epoch AI |
gpqa_diamond
decay ×0.26 |
0.92 ±0.02 | 1,467 ±8 | 1.5% | 2026-06-24 | max |
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 Coding 1,448 [1,444–1,451] Percentile 92 · #23 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
webdev/overall
|
1,605 ±3.24 | 1,480 ±1 | 39.5% | 2026-10-01 | max |
| LMArena |
text/coding
|
1,484 ±3.24 | 1,456 ±3 | 38.3% | 2026-10-02 | max |
| Epoch AI |
swe_bench_verified
decay ×0.37 |
0.79 ±0.02 | 1,477 ±10 | 13.9% | 2026-06-25 | max |
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 Agents 1,454 [1,449–1,459] Percentile 77 · #14 · components: 2
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
agent/overall
|
0.03 ±0.00 | 1,472 ±2 | 66.6% | 2026-10-02 | max |
| τ²-bench (Sierra) |
tau2/banking_knowledge
decay ×0.76 |
0.37 ±0.05 | 1,476 ±8 | 26.4% | 2026-07-24 | xhigh |
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 Math 1,442 [1,434–1,450] Percentile 89 · #29 · components: 3
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/math
|
1,484 ±6.76 | 1,474 ±6 | 51.5% | 2026-10-02 | max |
| Epoch AI |
frontiermath_t1_3
decay ×0.86 |
0.59 ±0.03 | 1,449 ±6 | 29.8% | 2026-06-19 | max |
| Epoch AI |
frontiermath_t4
decay ×0.86 |
0.29 ±0.07 | 1,439 ±12 | 11.1% | 2026-06-19 | max |
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 1,470 [1,465–1,475] Percentile 95 · #15 · components: 2 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/creative_writing
|
1,464 ±3.73 | 1,494 ±4 | 61.9% | 2026-10-02 | max |
| LMArena |
text/longer_query
|
1,480 ±2.82 | 1,475 ±3 | 32.0% | 2026-10-02 | max |
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 Language: 繁體中文 1,452 [1,444–1,460] Percentile 90 · #22 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/chinese
|
1,520 ±5.54 | 1,469 ±4 | 93.6% | 2026-10-02 | max |
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 Language: Deutsch 1,438 [1,422–1,456] Percentile 85 · #22 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/german
|
1,466 ±11 | 1,459 ±9 | 92.0% | 2026-10-02 | max |
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 Language: English 1,451 [1,445–1,456] Percentile 90 · #29 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/english
|
1,474 ±2.78 | 1,467 ±3 | 93.9% | 2026-10-02 | max |
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 Language: Español 1,450 [1,435–1,464] Percentile 93 · #11 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/spanish
|
1,477 ±8.74 | 1,470 ±8 | 92.5% | 2026-10-02 | max |
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 Language: Français 1,440 [1,426–1,454] Percentile 82 · #25 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/french
|
1,478 ±8.33 | 1,458 ±8 | 92.7% | 2026-10-02 | max |
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 Language: 日本語 1,444 [1,427–1,460] Percentile 86 · #19 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/japanese
|
1,450 ±12 | 1,464 ±9 | 92.1% | 2026-10-02 | max |
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 Language: 한국어 1,451 [1,436–1,466] Percentile 90 · #14 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/korean
|
1,447 ±11 | 1,471 ±8 | 92.5% | 2026-10-02 | max |
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 1,432 [1,415–1,449] Percentile 78 · #29 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/polish
|
1,453 ±11 | 1,452 ±9 | 92.0% | 2026-10-02 | max |
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 Language: Русский 1,450 [1,443–1,458] Percentile 88 · #28 · components: 1 provisional
| Source | Component | Value | Elo-eq | Weight | Measured | Variant |
|---|---|---|---|---|---|---|
| LMArena |
text/russian
|
1,467 ±4.45 | 1,467 ±4 | 93.7% | 2026-10-02 | max |
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 |
|---|---|
| GLM 5.2 | 1,442 A |
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.