The AI arena is free today

Open Superagent

DeepSeek-V4-Flash-0731 vs Hy4 preview

DeepSeek-V4-Flash-0731 and Hy4 preview are closely matched at 44.7 and 51.0 on the LLM Stats Score.

DeepSeek · Tencent · Updated for 2026

Which is better?

DeepSeek-V4-Flash-0731 and Hy4 preview are closely matched on the overall LLM Stats Score at 44.7 and 51.0.

In the 7 individual benchmarks reported for both models, Hy4 preview wins 6; this is a narrower head-to-head signal than the composite indexes.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose DeepSeek-V4-Flash-0731

  • you want predictable pricing at $0.06/M input and $0.18/M output

Choose Hy4 preview

  • you value its reported benchmark strengths — it wins 6 of 7 exact shared results
  • you want the most recent training data — it shipped Aug 2026

At a glance

The differences that matter most.

DeepSeek-V4-Flash-0731
DeepSeek
Hy4 preview
Tencent
Core performance indexes
LLM Stats Score
44.7
#36
51.0
#14
Reasoning
42.3
#46
50.8
#10
Coding
33.0
#37
39.7
#14
Agents
31.1
#31
36.7
#16
Cost, coverage & limits
Benchmark wins
1 of 7
6 of 7
Input price
$0.06 / M
— / M
Output price
$0.18 / M
— / M
Context window
1,048,576
—

Capability indexes

Additional strengths measured across groups of related public benchmarks

1 shared
Index
DeepSeek-V4-Flash-0731
Hy4 preview
Tool use
25.8#32
30.9#14
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

9 reported for DeepSeek-V4-Flash-0731 · 32 for Hy4 preview

7 shared

DeepSeek-V4-Flash-0731 outperforms in 1 benchmarks (Agents' Last Exam), while Hy4 preview is better at 6 benchmarks (AutomationBench, CyberGym, DeepSWE, NL2Repo, Terminal-Bench 2.1, Toolathlon).

Hy4 preview significantly outperforms across most benchmarks.

Sun Sep 20 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

466.0B diff

Hy4 preview has 466.0B more parameters than DeepSeek-V4-Flash-0731, making it 153.3% larger.

DeepSeek
DeepSeek-V4-Flash-0731
304.0Bparameters
Tencent
Hy4 preview
770.0Bparameters
304.0B
DeepSeek-V4-Flash-0731
770.0B
Hy4 preview

Context Window

Maximum input and output token capacity

Only DeepSeek-V4-Flash-0731 specifies input context (1,048,576 tokens). Only DeepSeek-V4-Flash-0731 specifies output context (1,048,576 tokens).

DeepSeek
DeepSeek-V4-Flash-0731
Input1,048,576 tokens
Output1,048,576 tokens
Tencent
Hy4 preview
Input- tokens
Output- tokens
Sun Sep 20 2026 • llm-stats.com

License

Usage and distribution terms

DeepSeek-V4-Flash-0731 is licensed under MIT, while Hy4 preview uses Apache 2.0.

License differences may affect how you can use these models in commercial or open-source projects.

DeepSeek-V4-Flash-0731

MIT

Open weights

Hy4 preview

Apache 2.0

Open weights

Release Timeline

When each model was launched

DeepSeek-V4-Flash-0731 was released on 2026-07-31, while Hy4 preview was released on 2026-08-28.

Hy4 preview is 1 month newer than DeepSeek-V4-Flash-0731.

DeepSeek-V4-Flash-0731

Jul 31, 2026

1 months ago

Hy4 preview

Aug 28, 2026

3 weeks ago

4w newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion→

Judge for yourself.

Run your own prompts against DeepSeek-V4-Flash-0731 and Hy4 preview side-by-side, then vote on the output you prefer.

DeepSeek-V4-Flash-0731
✓ Preferred
Hy4 preview
Open in Playground

FAQ

Common questions about DeepSeek-V4-Flash-0731 vs Hy4 preview.

Which is better, DeepSeek-V4-Flash-0731 or Hy4 preview?

DeepSeek-V4-Flash-0731 and Hy4 preview are closely matched on the LLM Stats Score at 44.7 and 51.0. DeepSeek-V4-Flash-0731 is made by DeepSeek and Hy4 preview is made by Tencent. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does DeepSeek-V4-Flash-0731 compare to Hy4 preview in benchmarks?

DeepSeek-V4-Flash-0731 scores Terminal-Bench 2.1: 82.7%, CyberGym: 76.7%, Toolathlon: 70.3%, DSBench-FullStack: 68.7%, DSBench-Hard: 59.6%. Hy4 preview scores GPQA: 92.3%, Terminal-Bench 2.1: 85.4%, WideSearch: 83.9%, MCP Atlas: 83.7%, SWE-bench Multilingual: 82.9%.

What are the context window sizes for DeepSeek-V4-Flash-0731 and Hy4 preview?

DeepSeek-V4-Flash-0731 supports 1.0M tokens and Hy4 preview supports an unknown number of tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between DeepSeek-V4-Flash-0731 and Hy4 preview?

Key differences include LLM Stats Score (44.7 vs 51.0), licensing (MIT vs Apache 2.0). See the full comparison above for benchmark-by-benchmark results.

Who makes DeepSeek-V4-Flash-0731 and Hy4 preview?

DeepSeek-V4-Flash-0731 is developed by DeepSeek and Hy4 preview is developed by Tencent.

Related comparisons