Across 12 comparable benchmarks, Qwen3.5-397B-A17B has better reported results on 12 and worse on 0.
Strong evidence12 of 12 shared results countedHow we compare
Price basis: Listed rates vs listed rates. The price section below shows the applicable periods.
| Fact | Qwen3-30B-A3B-2507 | Qwen3.5-397B-A17B |
|---|---|---|
| Context window | 256K tokens | 256K tokens |
| Input / 1M | — | $0.600 |
| Output / 1M | — | $3.60 |
| Blended (4:1) | — | $1.20 |
| Status | Deprecated | Available |
| Released | July 29, 2025 | February 16, 2026 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 6 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
Qwen3.5-397B-A17B ahead on 3 of 3 counted
Qwen3.5-397B-A17B ahead on 3 of 3 counted
Qwen3.5-397B-A17B ahead on 1 of 1 counted
Qwen3.5-397B-A17B ahead on 1 of 1 counted
Qwen3.5-397B-A17B ahead on 2 of 2 counted
No shared comparable results for Abstract reasoning, Mathematics, Computer & tool use, Long context, Cybersecurity.
Published scores, with matching modes where available. Gaps show Qwen3-30B-A3B-2507 relative to Qwen3.5-397B-A17B. Reading the results
| Benchmark | Qwen3-30B-A3B-2507 | Qwen3.5-397B-A17B | Gap |
|---|---|---|---|
| LiveCodeBenchAlgorithmic Coding · Pass @K | 43.2Normal · no tools | 83.6≠Thinking · no tools | 0.52×#92 vs #29 of 113 |
| SWE-bench VerifiedRepository Engineering · Accuracy | 22.0Thinking · no tools | 76.4≠Thinking · tools | −54.4 pts#107 vs #31 of 108 |
| Terminal Bench HardAgentic Development · Accuracy | 5.3Thinking · tools | 40.9Thinking · tools | −35.6 pts#135 vs #34 of 157 |
| CritPtScientific Reasoning · Score | 0.3Thinking · no tools | 1.7Thinking · no tools | 0.18×#124 vs #92 of 134 |
| GPQA DiamondScientific Reasoning · Accuracy | 70.4Normal · no tools | 88.4≠Thinking · no tools | −18.0 pts#138 vs #63 of 200 |
| SciCodeScientific Computing · Score | 33.0Thinking · no tools | 44.8Thinking · no tools | 0.74×#89 vs #64 of 98 |
| BrowseCompFact Finding · Accuracy | 2.3Thinking · tools | 78.6≠Thinking · tools + web | −76.3 pts#61 vs #25 of 61 |
| GDP.pdfDocuments & Charts · Task pass rate (%) | 0.4Thinking · no tools | 12.2Thinking · no tools | −11.8 pts#96 vs #52 of 96 |
| τ²-BenchService Workflows · Accuracy | 49.0Thinking · tools | 86.7Thinking · tools | −37.7 pts#33 vs #7 of 40 |
| τ³-BankingService Workflows · Score | 5.4Thinking · tools | 13.4Thinking · tools | 0.40×#107 vs #80 of 120 |
| HLEKnowledge Exams · Accuracy | 9.8Thinking · no tools | 48.3≠Thinking · tools + web | −38.5 pts#128 vs #56 of 151 |
| MMLU-ProKnowledge Exams · Accuracy | 78.4Normal · no tools | 87.8≠Thinking · no tools | −9.4 pts#69 vs #10 of 153 |
Algorithmic Coding · Pass @K
Repository Engineering · Accuracy
Agentic Development · Accuracy
Scientific Reasoning · Score
Scientific Reasoning · Accuracy
Scientific Computing · Score
Fact Finding · Accuracy
Documents & Charts · Task pass rate (%)
Service Workflows · Accuracy
Service Workflows · Score
A dash means no result is available.
| Benchmark | Qwen3-30B-A3B-2507 | Qwen3.5-397B-A17B |
|---|---|---|
| AIME 2026 | — | 91.3 |
| AIME2025 | 61.3 | — |
| C-Eval | — | 93.0 |
| FrontierMath v2 | — | 31.2 |
| IMO-AnswerBench | — | 80.9 |
| APEX-SWE | — | 25.1 |
| SWE-bench Multilingual | — | 69.3 |
| SWE-Bench Pro - Public | — | 50.9 |
| Terminal Bench 2.0 | — | 52.5 |
| APEX-Agents | — | 24.9 |
| OSWorld-Verified | — | 62.2 |
| Pinch Bench | — | 89.1 |
| Tool Decathlon | — | 38.3 |
| Toolathlon-Verified | — | 40.7 |
| ECI | — | 147 |
| Creative Writing | — | 1478 |
| IF Bench | — | 76.5 |
| Context Arena | — | 75.5 |
| LongBench v2 | — | 63.2 |
| MMMU | — | 85.0 |
| MMMU-Pro | — | 77.3 |
| Harvey Lab-AA | — | 72.4 |
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
A blended cost needs a published input and output price on both sides, and one of these two does not have them yet.
Qwen3-30B-A3B-2507: Listed text rates; pricing tier not confirmed.
Qwen3.5-397B-A17B: International pricing. context_length <= 256000 Alibaba Model Studio International (Singapore)
| Price / 1M tokens | Qwen3-30B-A3B-2507 | Qwen3.5-397B-A17B | Ratio |
|---|---|---|---|
| Input | — | $0.60 | — |
| Output | — | $3.60 | — |
| Cached input | — | — | — |
| Cache write | — | — | — |
| Blended | — | $1.20 | — |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
| Attribute | Qwen3-30B-A3B-2507 | Qwen3.5-397B-A17B |
|---|---|---|
| Availability | Deprecated | Available |
| Provider | Alibaba | Alibaba |
| Model type | Chat model | Multimodal model |
| First release | July 29, 2025 | February 16, 2026 |
| Context window | 256K tokens | 256K tokens |
| Max output | 16K tokens | — |
| Total parameters | 30.5B | 397B |
| Active parameters | 3.3B | 17B |
| Architecture | Mixture of Experts (MoE) | Mixture of Experts (MoE) |
| Documented capabilities | Multilingual | Reasoning, Multilingual |
| Input modalities | Text | Text, Image, Video |
| Output modalities | Text | Text |
| Weights | Open weights | Open weights |
| Licence | Apache 2.0 | Apache 2.0 |
| Commercial use | Commercial use allowed | Commercial use allowed |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used Qwen3-30B-A3B-2507 or Qwen3.5-397B-A17B? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.