whatmodel.sh / decision / Mercury Decide (free endpoint)
Mercury Decide (free endpoint)
↗ share
+ contribute
Mercury Decide (free endpoint) is a closed-weights decision model from Inception. Licence: proprietary.
Catalogue identity notes
System One Bench identifies the tested endpoint as inception/mercury-decide:free. Hosted service snapshot, not an immutable weight revision. Separate from chat or open-weight models with similar names. API access verified 2026-10-31; benchmark dates are retained per result. Results are for the :free endpoint; no equivalence to a differently routed paid endpoint is assumed.
$ how it compares
The decision models ranked either side of Mercury Decide (free endpoint), out of 357 with a score, then more from Inception.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 75 |  linne-fern 1.0 (27B) | 63±1 | model files | ||
| 76 |  Microsoft-Decision-1 | 63±1 | api price unknown | ||
| 77 | Mercury Decide (free endpoint) | 63±2 | api price unknown | ||
| 78 |  Nace.AI Drex v2.0 (35B-A3B) | 63±1 | model files | ||
| 79 |  Sieve-Vision-27B | 63±4 | model files | ||
| more from Inception | |||||
| 56 | Mercury Decide (paid endpoint) | 65±1 | api price unknown | ||
see all decision models · find the ones that fit your machine
$ what it's good at
retrieval & classification▁▁▁·▁▁▁▁▃▃█▅ #77 of 258 · 23% of its score · 6 tests
Beats AutoTrust-JEV-27B v1.1, oenogarum (Gemma 4 31B), Decider 31B (NVFP4) here, though they rank higher overall. Trails StartLux-Decision v1.1 0.8B, Standard One 3B SH (v4), Standard One 8B SH (v4) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| System One Bench: MIRACL pairwise (Arabic) | 90.0% ✎ | ▂···▁··▃▄▂█▇ beats 18/29 | 7.9% |
| System One Bench: MASSIVE intent (Arabic) | 85.0% ✎ | ▄·▁·▁▃▃▃█▄▇▇ beats 15/29 | 7.0% |
| System One Bench: MIRACL rerank (Arabic) | 88.0% ✎ | ▁▂·▁▁·▂·▄▄█▇ beats 17/29 | 4.5% |
| System One Bench: MASSIVE scenario (Arabic) | 72.0% ✎ | ▃▁·▄▃▂█▂▁▁·▃ beats 19/29 | 3.0% |
| System One Bench: AG News (English) | 92.0% ✎ | ▁···▅▁▅▂▆█▂▅ beats 28/29 | 0.8% |
| System One Bench: Banking77 (English) | 86.0% ✎ | █▅▆█▅▃▃▃···▂ beats 8/26 | 0.0% |
language understanding▂▂▂▃▂▅▆▃█▆▄▃ #50 of 256 · 21% of its score · 9 tests
Beats Ainvest-Decision-27B v1.0, Turion Jev 31B, MISA AI 2.0 Decision Plus · oak 1.1 (31B) here, though they rank higher overall. Trails Decider 35B-A3B, Sieve 9B v2, Clef-flash here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| System One Bench: BoolQ (English) | 90.5% ✎ | ▁▁▁▃·▁▃▃▇▇█▅ beats 19/29 | 5.4% |
| System One Bench: OSACT4 offensive (Arabic) | 92.0% ✎ | ▁······▁▃█▇▄ beats 21/29 | 4.9% |
| System One Bench: SST-2 (English) | 95.5% ✎ | ▁·▃▁▂▂▃▅█·▂▁ beats 22/29 | 3.3% |
| System One Bench: SST-5 (English) | 52.0% ✎ | ▃·▁▁▃·▇▄▄▄█▅ beats 9/29 | 3.1% |
| System One Bench: XNLI Arabic (Arabic) | 76.0% ✎ | ▄▁▃▁▄▇█▅▃▁·▃ beats 19/29 | 1.5% |
| System One Bench: XNLI French (French) | 82.7% ✎ | ▂▅·▁▁▅▅█▃▁·▃ beats 18/29 | 1.4% |
| System One Bench: Tweet sentiment (Arabic) | 62.5% ✎ | ▁······▁▁▄█▄ beats 6/29 | 1.2% |
| System One Bench: XNLI English (English) | 91.3% ✎ | ▃·▁·▃▇█▄█▃▃▁ beats 23/29 | 0.4% |
1 more test
| test | result | vs the others in its run | counts |
|---|---|---|---|
| System One Bench: DAIR Emotion (English) | 91.5% ✎ | ██▃·····▁··▃ beats 29/29 | 0.0% |
vision▁·▃▂▃▃▇▄▄█▄▃ #6 of 79 · 40% of its score · 5 tests
Beats Fastino GLiDE 1.2, StartLux-Decision v1.1 27B, vLLM-SR d3-ultra (Decision 3.0, 27B) here, though they rank higher overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| System One Bench: Image object count | 93.5% ✎ | ▃····▃█▃▅·█▃ beats 8/11 | 11% |
| System One Bench: SST-2 screenshot | 93.0% ✎ | █···███·█·▄▄ beats 3/11 | 8.7% |
| System One Bench: Visual yes/no questions | 87.0% ✎ | ▂······▃█▃·▃ beats 4/11 | 7.6% |
| System One Bench: Oxford pets classification | 84.5% ✎ | ███·█▄▄·▄··▄ beats 3/11 | 6.6% |
| System One Bench: Arabic sentiment screenshot | 39.0% ✎ | ▄··▄··▃···▆█ beats 4/11 | 6.2% |
workflows▁▁▁▂▁▂▂▆█▄▁▁ #34 of 114 * · 7.1% of its score · 1 test
Trails StartLux-Decision v1.1 2B, simplejev-qwen3.8-27b (DecisionBench 1d4bf0f2), xor (DecisionBench 679decd4) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| System One Bench: Typed decisions (English) | 71.1% ✎ | ▁▁·▁▁▁▅▄▇██▁ beats 16/29 | 7.1% |
other tests2 not in a task · 6.9% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Benchmark Heaven jevbench v1.6.1: intelligence (points) | 65.9070 points ✎ | ▃···▃·▃▃█▅▅█ beats 10/13 | 5.4% |
| Benchmark Heaven jevbench v1.6.1: calibration (points) | 81.3434 points ✎ | ▂······▂▆▂██ beats 3/13 | 1.5% |
$ where to run it
Through an API
api · OpenRouter · price unknown · openrouter.ai ↗
Model repository and documentation: openrouter.ai ↗
$ how the score was worked out · System One Bench (Ali Alhashi… 91%, Benchmark Heaven (own decisio… 7%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 System One Bench (Ali Alhashimi) | 21 | 91% |
 Benchmark Heaven (own decision evaluations) | 2 | 7% |
| prior · pull toward the middle | 2% | |
| weighted average | = 63.0 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json