EVALUATION
Benchmarks
Every model run by Blockway on one harness, with the same settings. v1 is ahead of Qwen3.8-27B on competitive coding and competition maths; Qwen3.8-27B still leads on agentic coding, multi-turn tool use, instruction following and GPQA.
| Benchmark | Agens Volundr 32B v1 | Volundr Preview | Qwen3.8-27B |
|---|
| LiveCodeBench v62025-02 to 2025-05, best of 2 runs | 67.9▲ +8.4 | 63.4 | 59.5 |
| HumanEvalgreedy completion | 78.7▲ +1.3 | 81.7 | 77.4 |
| SWE-bench Verified mini50 tasks, Codeway harness | 54.0 | 44.0 | 58.0 |
| BFCL v4 irrelevancenon-live, official harness | 82.1▲ +0.4 | 80.8 | 81.7 |
| τ²-benchairline, retail, telecom | 76.7 | 74.2 | 79.2 |
| BFCL v4 parallelnon-live, official harness | 94.0 | 92.0 | 94.0 |
| AIME 2025avg@4, best of 2 runs | 80.0▲ +6.7 | 76.7 | 73.3 |
| MATH-500 | 96.8▲ +0.2 | 98.2 | 96.6 |
| IFEvalprompt-strict | 87.2 | 87.6 | 88.4 |
| IFBenchprompt-strict | 64.3 | 66.3 | 69.3 |
| GPQA Diamondbest of 4 runs, 6K-token thinking budget | 80.3 | 84.3 | 85.4 |
| MMLU-Pro1,400-question subset | 80.0 | 79.9 | 80.4 |
| CMMLU2,010-question subset | 85.9 | 86.5 | 86.4 |
▲ v1 ahead of Qwen3.8-27B, in points. Best score in each row in bold.
Thinking on, temperature 0.6 and the same output-token limit for every model unless noted; answers cut off by the limit count as wrong. GPQA Diamond: 6K-token thinking budget, then the answer is forced, best of 4 runs for every model. τ²-bench: agent at temperature 0, Qwen3.8-27B as the user simulator for every model; the 13 retail tasks that need an LLM judge are scored as failed for every model. BFCL v4: non-live categories, official harness at temperature 0.001. SWE-bench Verified mini: a 50-task subset run through Codeway, Blockway's coding-agent harness, with each model's default sampling and identical limits. Qwen3.8-27B was run by us on the same harness; these are not its publisher's figures. Run by run (best shown): GPQA Diamond v1 79.3 / 78.8 / 80.3 / 78.8, Preview 79.3 / 79.8 / 83.3 / 84.3, Qwen3.8-27B 83.8 / 84.8 / 82.3 / 85.4; AIME 2025 v1 78.3 / 80.0, Preview 76.7 / 72.5, Qwen3.8-27B 70.0 / 73.3; LiveCodeBench v1 66.4 / 67.9, Preview 63.4 / 63.4, Qwen3.8-27B 59.5 / 58.8.
Smaller builds
The INT4 and 16 GB builds against BF16 v1. The 16 GB build trades some accuracy for size, mostly on hard reasoning and code.
| Benchmark | BF16 | INT4 | 16 GB | Setting |
|---|
| AIME 2025 | 80.0 | — | 80.0 | avg@4, best of 2 runs (bf16 mean 79.2); 16 GB: best of 4 (mean 76.7) |
| MATH-500 | 96.8 | — | 96.4 | |
| GPQA Diamond | 80.3 | 81.3 | 73.7 | best of 4 runs, 6K-token thinking budget then a forced answer; INT4 mean 78.9 |
| HumanEval | 78.7 | 73.8 | 73.2 | greedy completion |
| IFEval | 87.2 | — | 86.5 | prompt-strict |
| BFCL v4 parallel | 94.0 | — | 94.0 | non-live, official harness |
| BFCL v4 irrelevance | 82.1 | — | 81.7 | 16 GB: provisional, 8 of 240 hit a 32K client cap |
16 GB build vs BF16: mean KL 0.117, top-1 agreement 90.4%. Full tables, settings and known limitations are on the model cards.