Model comparison
Phi-4 vs Qwen3.5 397B
Head-to-head evidence from 11 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: Phi-4 #202 (Supported); Qwen3.5 397B #77 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Phi-4 and Qwen3.5 397B share 11 comparable benchmark results. 0 of 8 categories are comparable. 0 results are unique to Phi-4; 44 to Qwen3.5 397B.
Updated July 24, 2026- Shared results
- 11
- Phi-4 only
- 0
- Qwen3.5 397B only
- 44
- Comparable categories
- 0 / 8
Benchmark data for Phi-4 and Qwen3.5 397B is coming soon on BenchLM.
Confidence note. This is a partial-evidence comparison with 11 shared benchmark results across 5 evidence categories; 0 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
BenchLM has partial data for these models, but not enough overlapping benchmark coverage to produce a fair score-level comparison yet.
Qwen3.5 397B is priced at $0.60 input / $3.60 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Phi-4. Qwen3.5 397B has the larger context window at 128K, compared with 16K for Phi-4.
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 | Phi-4 | Δ | Qwen3.5 397B |
|---|---|---|---|
| Agentic | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B56.5 |
| Coding | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B66.5 |
| Reasoning | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B63.2 |
| Knowledge | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B56.6 |
| Math | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B90.6 |
| Multilingual | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B84.7 |
| Multimodal | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B79.6 |
| Inst. Following | Phi-4Not measured | MarginNo overlap | Qwen3.5 397B92.6 |
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Phi-4 | Qwen3.5 397B | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Phi-4$0 input / $0 output | Qwen3.5 397B$0.6 input / $3.6 output | Phi-4 has the lower combined listed price. |
| Generation speedtokens per second | Phi-435 tok/s | Qwen3.5 397B96 tok/s | Qwen3.5 397B has the higher measured throughput. |
| First-answer latencyseconds to first token | Phi-42.02 s | Qwen3.5 397B2.44 s | Phi-4 reaches the first token sooner. |
| Context windowmaximum listed tokens | Phi-416K | Qwen3.5 397B128K | Qwen3.5 397B lists the larger context window. |
Benchmark Deep Dive
Agentic18 benchmarks
| Benchmark | Phi-4 | Qwen3.5 397B | Result |
|---|---|---|---|
| τ²-bench resultsSource | 0% | 95.6% | Qwen3.5 397B leads |
| Terminal-Bench 2.0Source | — | 52.5% | Not comparable |
| BrowseCompSource | — | 62% | Not comparable |
| Claw-EvalSource | — | 56.8% | Not comparable |
| QwenClawBenchSource | — | 51.8% | Not comparable |
| τ³-bench resultsSource | — | 68.4% | Not comparable |
| VITA-BenchSource | — | 43.7% | Not comparable |
| DeepPlanningSource | — | 37.6% | Not comparable |
| ToolathlonSource | — | 36.3% | Not comparable |
| MCP AtlasSource | — | 46.1% | Not comparable |
| MCP-TasksSource | — | 74.2% | Not comparable |
| WideResearchSource | — | 74.0% | Not comparable |
| Gert LabsSource | — | 46.76% | Not comparable |
| ResearchClawBenchSource | — | 14.2% | Not comparable |
| AA Agentic IndexSource | — | 19.9% | Not comparable |
| APEX-Agents-AASource | — | 15.3% | Not comparable |
| GDPval-AASource | — | 23.1% | Not comparable |
| GDPval-AASource | — | 962 | Not comparable |
Coding5 benchmarks
Reasoning4 benchmarks
Knowledge12 benchmarks
| Benchmark | Phi-4 | Qwen3.5 397B | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 4.9% | 33.7% | Qwen3.5 397B leads |
| AA-GPQA DiamondSource | 57.5% | 89.3% | Qwen3.5 397B leads |
| AA-HLESource | 4.1% | 27.3% | Qwen3.5 397B leads |
| AA-Omniscience IndexSource | -56.7% | -29.8% | Qwen3.5 397B leads |
| AA-Omniscience AccuracySource | 13.2% | 31.4% | Qwen3.5 397B leads |
| AA-Omniscience Hallucination RateSource | 80.5% | 89.1% | Phi-4 leads |
| GPQASource | — | 88.4% | Not comparable |
| SuperGPQASource | — | 70.4% | Not comparable |
| MMLU-ProSource | — | 87.8% | Not comparable |
| MMLU-ReduxSource | — | 94.9% | Not comparable |
| C-EvalSource | — | 93% | Not comparable |
| HLESource | — | 28.7% | Not comparable |
Math5 benchmarks
Multilingual2 benchmarks
Multimodal7 benchmarks
Frequently Asked Questions (3)
Can I compare Phi-4 and Qwen3.5 397B on BenchLM yet?
Not fully yet. BenchLM is tracking both models, but the sourced benchmark breakdown for this comparison is still coming soon.
Why does this comparison show “coming soon”?
BenchLM only shows category winners and benchmark-level calls when we have sourced results that can be compared fairly. For these models, the public benchmark coverage is not complete enough yet.
What data is available for Phi-4 and Qwen3.5 397B today?
Phi-4: $0.00 input / $0.00 output per 1M tokens Qwen3.5 397B: $0.60 input / $3.60 output per 1M tokens Both model pages still include creator, context window, reasoning mode, and other metadata while benchmark coverage fills in.