# PostTrainBench (lab runs) leaderboard

Coding benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelchoice.dev/leaderboards/benchmarks/coding-post-train-bench

PostTrainBench lab-run results with no stated version. Every model the catalog carries a published PostTrainBench (lab runs) value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Coding |
| Measure | Weighted aggregate as reported by each lab |
| Tasks | Post-training base models under the PostTrainBench budget |
| Difficulty | Autonomous ML engineering |
| Published by | [PostTrainBench](https://posttrainbench.com/) |

PostTrainBench results that model makers ran themselves and published without naming the benchmark version.

## PostTrainBench (lab runs) ranking

3 models with a published PostTrainBench (lab runs) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Weighted aggregate as reported by each lab | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [GLM-5.3](https://modelchoice.dev/models/glm-5-3) | Z.AI | 39.8 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Kimi K3](https://modelchoice.dev/models/kimi-3) | Moonshot AI | 36.6 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Hy4 preview](https://modelchoice.dev/models/hy4-preview) | Tencent | 35.6 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
