Part of Best Decision Model for Agent Routing and Scoring
Best Open-Weight Decision Model for Self-Hosting
swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) is the provisional first pick for Best Open-Weight Decision Model for Self-Hosting as of November 15, 2026, while independent task measurements are still limited.[1][2] It scores 72.47 on JevBench (#11 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 56 and its calibration axis 89, each out of 100. Winnow-12B Q8 scores within two points, so read the top two as close rather than as a measured gap. Winnow-12B Q8 is the next-ranked alternative. It scores 72.21 on JevBench (#12 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 60 and its calibration axis 83, each out of 100.
About this recommendation
- Updated
- Evidence through
- Sources
- 8
- Revision
- v48
Comparing Jev alternatives or choosing a System One model? Read our guide to AI decision models, routing, classification and scoring for the decision model vs LLM tradeoffs and a self-hosting checklist.
Decision audit
Why this result
Inspect the inputs and the computed order behind the recommendation.
Models screened
24
live candidates
Evaluation feeds
1
task-weighted
Winner coverage
100%
intended feed weight
Largest provider share
2 of 8
blockbrain
swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) leads the measured task comparison with limited independent task measurements. Community volume, release age and usage popularity add no ranking points. The leading scores are close under this policy.
Real-world track record is still developing. The recommendation can lead on independent measurements while practitioner reports accumulate.
Sources evaluated
These are the weights used in scoring. Missing measurements are omitted from each model’s average. A feed with no result for any model released in the last 60 days is stale and does not score until it catches up.
| Evaluation feed | Weight | Winner result | Field measured |
|---|---|---|---|
| JevBenchObserved 2026-11-15 · evaluation date not supplied · swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) | 100% | #11 | 24/24 |
Provider concentration
Each exact model is scored separately; provider identity is not a ranking input.
- blockbrain2 models
- igor-l.-ninfer-contributors2 models
- akhilaaa31 model
- eldanring1 model
- mapika1 model
- razorback161 model
Decision table
Every published model is shown in computed order. Practitioner sources are distinct community threads, not the citations repeated in the prose below.
blockbrain · coverage 100% · provisional evidence
#11 JevBench
eldanring · coverage 100% · close score · provisional evidence
#12 JevBench
blockbrain · coverage 100% · close score · provisional evidence
#14 JevBench
akhilaaa3 · coverage 100% · provisional evidence
#18 JevBench
razorback16 · coverage 100% · provisional evidence
#28 JevBench
igor-l.-ninfer-contributors · coverage 100% · provisional evidence
#31 JevBench
igor-l.-ninfer-contributors · coverage 100% · provisional evidence
#32 JevBench
mapika · coverage 100% · provisional evidence
#33 JevBench
Task score averages the available, task-weighted measurements. Benchmark positions compare models matched to our catalog. Models measured by at least two independent evaluator families are considered first. Missing tests, practitioner report counts, popularity and release age add no points. Coverage describes evidence breadth, not a probability of correctness. Scores within two points are shown as close under this policy; this is not a statistical confidence interval. When the leading scores are that close, the model measured by more evaluator families ranks first.
- 01
swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4)blockbrain · open-weight
swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) scores 72.47 on JevBench (#11 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 56 and its calibration axis 89, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 02
Winnow-12B Q8eldanring · open-weight
Winnow-12B Q8 scores 72.21 on JevBench (#12 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 60 and its calibration axis 83, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 03
Cygnet (blockbrain, frozen Gemma-4-12B-it)blockbrain · open-weight
Gemma-4-12B-based variant with competitive calibration (87) and mid-tier composite performance (71.88), viable for teams already running Gemma infrastructure.
Best when: Rely on strong calibration (87) for scoring tasks where probability calibration drives downstream business logic.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 04
Jev-Omni (akhilaaa3, Gemma-4-12B merged)akhilaaa3 · open-weight
Jev-Omni (akhilaaa3, Gemma-4-12B merged) scores 70.99 on JevBench (#18 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 56 and its calibration axis 87, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 05
OpenJev (thinking, BF16)razorback16 · open-weight
OpenJev (thinking, BF16) scores 68.4 on JevBench (#28 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 70 and its calibration axis 81, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 06
NInfer Qwen3.8-Flash-Next mixedigor-l.-ninfer-contributors · open-weight
NInfer Qwen3.8-Flash-Next mixed scores 67.76 on JevBench (#31 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 49 and its calibration axis 90, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 07
NInfer Qwen3.8-27B NVFP4igor-l.-ninfer-contributors · open-weight
NInfer Qwen3.8-27B NVFP4 scores 66.69 on JevBench (#32 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 52 and its calibration axis 84, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
- 08
decider-35b-a3b (Mapika)mapika · open-weight
decider-35b-a3b (Mapika) scores 66.51 on JevBench (#33 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 49 and its calibration axis 84, each out of 100.
Best when: Its strongest result for this task is on JevBench, which measures calibrated typed decisions.No practitioner reports are linked yet, so no cautions are listed. Absence of reports is not evidence of reliability.
Frequently asked
- What is the best model for Best Open-Weight Decision Model for Self-Hosting?
- swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) ranks first as of November 15, 2026. swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) scores 72.47 on JevBench (#11 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 56 and its calibration axis 89, each out of 100.[1]
- What is the best alternative to swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4)?
- Winnow-12B Q8 ranks second. Winnow-12B Q8 scores 72.21 on JevBench (#12 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 60 and its calibration axis 83, each out of 100.[2]
- Can I self-host the top pick for Best Open-Weight Decision Model for Self-Hosting?
- Yes. swanOne (blockbrain, Qwen3.8-Flash-Next NVFP4) is open-weight, so it can run on your own hardware as well as through hosted routes.[1]
Sources
- 1
“Within the 126 catalog models matched to this evaluation: Scores 72.47 on JevBench (#11 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 56 and its calibration axis 89, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 2
“Within the 126 catalog models matched to this evaluation: Scores 72.21 on JevBench (#12 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 60 and its calibration axis 83, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 3
“Within the 126 catalog models matched to this evaluation: Scores 71.88 on JevBench (#14 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 55 and its calibration axis 87, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 4
“Within the 126 catalog models matched to this evaluation: Scores 70.99 on JevBench (#18 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 56 and its calibration axis 87, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 5
“Within the 126 catalog models matched to this evaluation: Scores 68.4 on JevBench (#28 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 70 and its calibration axis 81, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 6
“Within the 126 catalog models matched to this evaluation: Scores 67.76 on JevBench (#31 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 49 and its calibration axis 90, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 7
“Within the 126 catalog models matched to this evaluation: Scores 66.69 on JevBench (#32 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 52 and its calibration axis 84, each out of 100.”
JevBench · Benchmark · Nov 15, 2026 - 8
“Within the 126 catalog models matched to this evaluation: Scores 66.51 on JevBench (#33 of 126), Benchmark Heaven's composite of decision accuracy, calibration, speed, and cost for typed decision models; its decision accuracy axis is 49 and its calibration axis 84, each out of 100.”
JevBench · Benchmark · Nov 15, 2026
Rankings synthesized from community evidence and open benchmarks. See methodology. Not driven by vendor marketing.