Benchmarks · Professional work · Agents & computer use
Corpo Index
Star014 consulting-style tasks, each ending in a finished deliverable such as a note, minutes, a workbook or a deck, marked out of 10 by two blind graders. The mean of two roles: the model working alone, and the model writing one work order for a fixed GPT-6 Luna worker. Each model runs in its own harness. From local-bench.ai.
Quality 68 of 100 · Private test set · Has measured 10 of today's 10 default lineup models
23 models · 23 results · updated 2026-11-19 · local-bench.ai ↗
Leaderboard
Benchmark quality
How the score works- Overall68
68 of 100
How useful it is for telling today’s models apart
- Headroom20%29
Top models average 68 of 100
Can top models still improve on it?
- Test set20%100
Private
Could labs train on the questions?
- Recency15%100
36 days old
How new is it?
- Grading15%65
Not researched
How are answers scored?
- New models15%100
10 of 10 top models
How fast does it test new releases?
- Efforts10%0
One effort per model
Does it test each model at several reasoning efforts?
- Breadth5%57
23 models
How many models has it measured?
How each part was worked out
- Headroom 29 of 100 · 20% · +5.8
- From the spread of published scores and the room left above the top quarter of models, one score per model.
- Test set 100 of 100 · 20% · +20
- Private test set. The 14 consulting-style tasks were built for the benchmark and are not published; the method page describes them but shows none. Source
- Recency 100 of 100 · 15% · +15
- Released 2026-10-17 (from the publisher's pages, not yet checked by hand).
- Grading 65 of 100 · 15% · +9.8
- Judged from its unit and description; the grading method has not been researched.
- New models 100 of 100 · 15% · +15
- Has measured 10 of today's 10 default lineup models. Too few new models to time how quickly it adds them.
- Efforts 0 of 100 · 10% · 0
- Measured 0 of the 23 models that come in more than one reasoning effort at two or more of them.
- Breadth 57 of 100 · 5% · +2.9
- 23 models measured.
Not counted in the score: 0 stars · used by 1 published page. Both lift it in the directory’s recommended order. Test set, release date and grading were researched from the publisher’s pages (2026-11-19); tell us if one is wrong.
11 more models under All models in settings.
More from this benchmark
1 published measuresScore vs any published measure
11 more models under All models in settings.
All results
23 rows| # | Model | Effort | Corpo Index | Subscription $ / task | API $ / task |
|---|---|---|---|---|---|
| 1 | Opus 5.5 | xhigh | 70.6 ±5.1 | $0.023 | $1.80 |
| 2 | GPT-6.2 Astra | xhigh | 70.1 ±6.2 | $0.051 | $0.76 |
| 3 | GPT-6.2 Sol | xhigh | 67.2 ±5.8 | $0.008 | $0.24 |
| 4 | Fable 5.2 | xhigh | 66.9 ±6.8 | $0.19 | $5.18 |
| 5 | Sonnet 5.5 | xhigh | 66.3 ±5.7 | $0.012 | $0.89 |
| 6 | GPT-6.1 Sol | xhigh | 66.0 ±6.3 | $0.009 | $0.28 |
| 7 | GPT-6 Astra | xhigh | 65.4 ±6.7 | $0.047 | $1.32 |
| 8 | GPT-6 Sol | xhigh | 63.2 ±5.3 | $0.009 | $0.28 |
| 9 | Gemini 4 Carbon | high | 63.0 ±5.5 | $0.15 | $2.54 |
| 10 | Gemini 4 Neon | xhigh | 62.5 ±6.5 | $0.035 | $1.66 |
| 11 | Fable 5.1 | xhigh | 62.4 ±5.4 | $0.52 | $3.35 |
| 12 | Gemini 4 Hydrogen | xhigh | 61.3 ±4.9 | $0.016 | $0.77 |
| 13 | Haiku 5.5 | xhigh | 61.0 ±6.4 | $0.004 | $0.22 |
| 14 | GPT-6.1 Terra | xhigh | 60.1 ±5.4 | $0.002 | $0.060 |
| 15 | DeepSeek V4.1 Flash | high | 59.4 ±5.8 | $0.012 | $0.12 |
| 16 | GPT-6 Luna | xhigh | 56.2 ±4.6 | $0.001 | $0.020 |
| 17 | Hy4 preview | high | 55.1 ±4.5 | $0.13 | $0.38 |
| 18 | GLM 5.3 Flash | high | 54.8 ±5.0 | $0.007 | $0.040 |
| 19 | Muse Ember | xhigh | 53.9 ±4.7 | $0.040 | $0.040 |
| 20 | MiniMax M3.1 | xhigh | 53.7 ±5.0 | $0.045 | $0.090 |
| 21 | Hy4 Flash | high | 52.4 ±4.7 | $0.010 | $0.010 |
| 22 | Sonnet 5 | xhigh | 50.9 ±5.2 | $0.010 | $0.98 |
| 23 | Opus 4.6 | max | 49.1 ±5.6 | $1.26 | $1.26 |