# Hy4 preview vs Mellum2.1-12B-A2.5B-Thinking

> Side-by-side benchmark comparison for Hy4 preview and Mellum2.1-12B-A2.5B-Thinking across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/hy4-preview-vs-mellum2-1-12b-a2-5b-thinking
- Last updated: October 8, 2026

- Shared sourced benchmarks: 3
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

Evidence is too thin for a benchmark verdict: Hy4 preview and Mellum2.1-12B-A2.5B-Thinking share only 3 sourced benchmark rows.

## Summary

- Only 3 shared sourced benchmark rows are available. That is not enough for a broad capability verdict, so this mirror does not name an overall winner.
- Hy4 preview has the larger published context window at 1M.

## Model Snapshot

| Property | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking |
|----------|----------|----------|
| Creator | Tencent | JetBrains |
| Type | Open Weight | Open Weight |
| Reasoning | Reasoning | Reasoning |
| Context | 1M | 128K |
| Independent public score (not head-to-head) | 60.91 | null |
| Benchmarks Covered | 30 | 7 |
| Pricing (input/output) | $0.00 / $0.00 | $0.00 / $0.00 |

## Category Breakdown

### Agentic

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: 54.5 (Estimated · #39/122)
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.1 | 85.4% | 17.4% | Hy4 preview |
| CyberGym | 78.4% | Coming soon | Coming soon |
| WideResearch | 83.9% | Coming soon | Coming soon |
| DRACO | 77.2% | Coming soon | Coming soon |
| MCP Atlas | 83.7% | Coming soon | Coming soon |
| Toolathlon-Verified | 74.1% | Coming soon | Coming soon |
| APEX-Agents | 37.1% | Coming soon | Coming soon |
| skillsBench | 62.9% | Coming soon | Coming soon |
| JobBench | 61.7% | Coming soon | Coming soon |
| Agents' Last Exam | 22.8% | Coming soon | Coming soon |
| AutomationBench | 32.1% | Coming soon | Coming soon |
| BankerToolBench | 78.6% | Coming soon | Coming soon |
| HLE w/ tools | 55.4% | Coming soon | Coming soon |
| Terminal-Bench 2.1 (Vals) | 55.1% | Coming soon | Coming soon |
| CWE-bench v1 | 53.0% | Coming soon | Coming soon |

### Coding

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: 57.7 (Supported · #25/146)
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.1 | 85.4% | Coming soon | Coming soon |
| SWE-bench Pro | 65.7% | 28% | Hy4 preview |
| SWE Multilingual | 82.9% | Coming soon | Coming soon |
| DeepSWE | 64.3% | Coming soon | Coming soon |
| NL2Repo | 58.9% | Coming soon | Coming soon |
| ProgramBench | 17.5% | Coming soon | Coming soon |
| PostTrain Bench | 35.6% | Coming soon | Coming soon |
| sweMarathon | 31.9% | Coming soon | Coming soon |
| LiveCodeBench v6 | Coming soon | 82.0% | Coming soon |
| SWE-bench Verified | Coming soon | 47% | Coming soon |

### Reasoning

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: Coming soon
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| CritPt | 16.9% | Coming soon | Coming soon |

### Multimodal & Grounded

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: 84.9 (Unranked · 1 rankable row)
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| OfficeQA Pro | 66.2% | Coming soon | Coming soon |

### Knowledge

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: 62.7 (Estimated · #38/174)
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 92.3% | Coming soon | Coming soon |
| GPQA-D | 92.3% | 64.6% | Hy4 preview |
| HLE | 55.4% | Coming soon | Coming soon |
| HLE w/o tools | 43.4% | Coming soon | Coming soon |
| MMLU-Redux | Coming soon | 87.8% | Coming soon |

### Instruction Following

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: Coming soon
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| IFEval | Coming soon | 90.6% | Coming soon |

### Mathematics

- Winner: Insufficient shared evidence
- Hy4 preview public-lane score: Coming soon
- Mellum2.1-12B-A2.5B-Thinking public-lane score: Coming soon

| Benchmark | Hy4 preview | Mellum2.1-12B-A2.5B-Thinking | Winner |
|-----------|-----------|-----------|--------|
| Apex | 74.2% | Coming soon | Coming soon |

## FAQ

### Is there enough evidence to pick between Hy4 preview and Mellum2.1-12B-A2.5B-Thinking?

Not yet. The models share only 3 sourced benchmark rows, so this mirror reports the available measurements and metadata without naming a broad winner.

### What can I compare safely on this page?

Published pricing, context windows, model type, and individual shared benchmark rows are directly comparable. Independent overall scores and non-overlapping benchmark coverage are labeled separately and should not be read as a controlled head-to-head result.

## Related Comparisons

- [Hy4 preview vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-hy4-preview)
- [Mellum2.1-12B-A2.5B-Thinking vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-mellum2-1-12b-a2-5b-thinking)
- [Hy4 preview vs GPT-6 Astra](/compare/gpt-6-astra-vs-hy4-preview)
- [Mellum2.1-12B-A2.5B-Thinking vs GPT-6 Astra](/compare/gpt-6-astra-vs-mellum2-1-12b-a2-5b-thinking)

## Explore More

- [Hy4 preview profile](/models/hy4-preview)
- [Mellum2.1-12B-A2.5B-Thinking profile](/models/mellum2-1-12b-a2-5b-thinking)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)