Benchmark scores
Higher is better for all benchmark scores (SWE-bench, GPQA, HLE, Terminal-Bench, etc.). Green marks the model with the higher score.
LLM Comparison
Hy3 vs DeepSeek V4 Pro: benchmark scores, pricing & comparison.
Side-by-side Hy3 vs DeepSeek V4 Pro comparison across SWE-bench, GPQA, HLE, Terminal-Bench, coding agent scores, token pricing, context window, and AskClash RWT. Green marks the winner on each benchmark.
Rank #18 vs #53AskClash overall scores 58.7 vs 20.0.
Pricing $0.12/$0.43 vs $1.74/$3.48Input and output token prices per 1M tokens when published.
Open Weight vs Open WeightTencent vs DeepSeek.
Hy3 vs DeepSeek V4 Pro benchmark comparison
Green cells highlight the winning model for each metric. Scores are cached from the AskClash LLM leaderboard snapshot.
| Metric | Hy3 | DeepSeek V4 Pro |
|---|---|---|
| Overall Score | 58.7 | 20.0 |
| Leaderboard Rank | #18 | #53 |
| RWT | 8.0 | 7.0 |
| HLE | 53.2 | 7.7 |
| GPQA | 90.4 | 72.9 |
| MATH-500 | — | 64.5 |
| SWE-bench | 78.0 | 73.6 |
| SWE-Pro | 57.9 | 52.1 |
| Terminal-Bench | 71.7 | 59.1 |
| MCP Atlas | 79.1 | 69.4 |
| Finance Agent | — | 44.1 |
| Tau2 | 92.7 | — |
| MRCR | 75.1 | 44.7 |
| Input Price (per 1M tokens) | $0.12 | $1.74 |
| Output Price (per 1M tokens) | $0.43 | $3.48 |
| Context Window | 256K | 1M |
| Benchmarks Published | 9 | 8 |
Hy3 vs DeepSeek V4 Pro head-to-head charts
Hy3 leads 9 and DeepSeek V4 Pro leads 0 of 9 shared benchmarks. Hy3 is cheaper on a defined 1M-input / 200K-output workload. Charts show only benchmarks both models publish.
Hy3DeepSeek V4 Pro
Overall
RWT
HLE
GPQA
SWE-bench
SWE-Pro
Terminal-Bench
MCP Atlas
MRCR
Hy3
Input$0.12
Output$0.43
Workload$0.21
Context256K
DeepSeek V4 Pro
Input$1.74
Output$3.48
Workload$2.44
Context1M
Workload = published cost of 1M input + 200K output tokens. Open the live leaderboard for interactive compare charts.
More Hy3 and DeepSeek V4 Pro comparisons
Explore how Hy3 and DeepSeek V4 Pro stack up against other top-ranked LLMs.
How to read this comparison
Token pricing
Lower is better for input and output prices. Green marks the cheaper model per 1M tokens.
Coverage matters
Models with fewer disclosed benchmark cells may have inflated percentile scores. Check the benchmark cell count for context.
This comparison page is generated from the AskClash LLM leaderboard cache. Open the live leaderboard for real-time scores and interactive filtering.