Model comparison
DeepSeek V3.2 vs Step 3.7 Flash
Head-to-head evidence from 15 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
BenchAlign evidence: DeepSeek V3.2 supported; Step 3.7 Flash estimated. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. DeepSeek V3.2 and Step 3.7 Flash share 15 comparable benchmark results. 1 of 8 categories are comparable. 5 results are unique to DeepSeek V3.2; 15 to Step 3.7 Flash.
Updated July 16, 2026- Shared results
- 15
- DeepSeek V3.2 only
- 5
- Step 3.7 Flash only
- 15
- Comparable categories
- 1 / 8
Pick Step 3.7 Flash if you want the stronger benchmark profile. DeepSeek V3.2 only becomes the better choice if coding is the priority or you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 15 shared benchmark results across 6 evidence categories; 1 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
Step 3.7 Flash has the cleaner provisional overall profile here, landing at 57 versus 54. It is a real lead, but still close enough that category-level strengths matter more than the headline number.
Step 3.7 Flash is also the more expensive model on tokens at $0.20 input / $1.15 output per 1M tokens, versus $0.28 input / $0.42 output per 1M tokens for DeepSeek V3.2. That is roughly 2.7x on output cost alone. Step 3.7 Flash is the reasoning model in the pair, while DeepSeek V3.2 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Step 3.7 Flash gives you the larger context window at 256K, compared with 128K for DeepSeek V3.2.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | DeepSeek V3.2 | Δ | Step 3.7 Flash |
|---|---|---|---|
| Coding | DeepSeek V3.260.9 | Margin← 4.6 | Step 3.7 Flash56.3 |
| Agentic | DeepSeek V3.2Not measured | MarginNo overlap | Step 3.7 Flash66.4 |
| Math | DeepSeek V3.217.1 | MarginNo overlap | Step 3.7 FlashNot measured |
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | DeepSeek V3.2 | Step 3.7 Flash | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | DeepSeek V3.2$0.28 input / $0.42 output | Step 3.7 Flash$0.2 input / $1.15 output | DeepSeek V3.2 has the lower combined listed price. |
| Generation speedtokens per second | DeepSeek V3.235 tok/s | Step 3.7 FlashNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | DeepSeek V3.23.75 s | Step 3.7 FlashNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | DeepSeek V3.2128K | Step 3.7 Flash256K | Step 3.7 Flash lists the larger context window. |
Benchmark Deep Dive
Agentic13 benchmarks
| Benchmark | DeepSeek V3.2 | Step 3.7 Flash | Result |
|---|---|---|---|
| Claw-EvalSource | 40.2% | 67.1% | Step 3.7 Flash leads |
| VITA-BenchSource | 18.5% | — | Not comparable |
| τ²-bench resultsSource | 78.9% | 98.5% | Step 3.7 Flash leads |
| Gert LabsSource | 29.57% | 51.57% | Step 3.7 Flash leads |
| Terminal-Bench 2.0Source | — | 59.5% | Not comparable |
| BrowseCompSource | — | 75.8% | Not comparable |
| DeepSearchQASource | — | 92.8% | Not comparable |
| GDPval-AASource | — | 25.9% | Not comparable |
| ToolathlonSource | — | 49.5% | Not comparable |
| HLE w/ toolsSource | — | 47.2% | Not comparable |
| AA Agentic IndexSource | — | 21.5% | Not comparable |
| GDPval-AASource | — | 1017 | Not comparable |
| APEX-Agents-AASource | — | 14.8% | Not comparable |
CodingDeepSeek V3.2 wins7 benchmarks
| Benchmark | DeepSeek V3.2 | Step 3.7 Flash | Result |
|---|---|---|---|
| SWE-RebenchSource | 60.9% | — | Not comparable |
| React Native EvalsSource | 71.5% | — | Not comparable |
| Terminal-Bench HardSource | 32.6% | 35.6% | Step 3.7 Flash leads |
| AA-SciCodeSource | 38.7% | 40.0% | Step 3.7 Flash leads |
| SWE-bench ProSource | — | 56.3% | Not comparable |
| Terminal-Bench 2.0Source | — | 59.5% | Not comparable |
| AA Coding IndexSource | — | 39.6% | Not comparable |
Reasoning2 benchmarks
Knowledge6 benchmarks
| Benchmark | DeepSeek V3.2 | Step 3.7 Flash | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 24.7% | 30.3% | Step 3.7 Flash leads |
| AA-GPQA DiamondSource | 75.1% | 80.9% | Step 3.7 Flash leads |
| AA-HLESource | 10.5% | 19.9% | Step 3.7 Flash leads |
| AA-Omniscience IndexSource | -46.7% | -37.5% | Step 3.7 Flash leads |
| AA-Omniscience AccuracySource | 24.2% | 25.4% | Step 3.7 Flash leads |
| AA-Omniscience Hallucination RateSource | 93.5% | 84.4% | Step 3.7 Flash leads |
Math2 benchmarks
Multimodal4 benchmarks
Inst. Following1 benchmarks
| Benchmark | DeepSeek V3.2 | Step 3.7 Flash | Result |
|---|---|---|---|
| AA-IFBenchSource | 49.0% | 67.3% | Step 3.7 Flash leads |
Frequently Asked Questions (2)
Which is better, DeepSeek V3.2 or Step 3.7 Flash?
Step 3.7 Flash is ahead on BenchLM's provisional leaderboard, 57 to 54.
Which is better for coding, DeepSeek V3.2 or Step 3.7 Flash?
DeepSeek V3.2 has the edge for coding in this comparison, averaging 60.9 versus 56.3. Inside this category, Terminal-Bench Hard is the benchmark that creates the most daylight between them.
Related Comparisons
Explore More
The AI models change fast. We track them for you.
A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.