StepFun: Step 3.7 Flash vs Google: Gemma 4 31B
Side-by-side comparison of specs, pricing, benchmark scores, and task rankings. Updated 2026-11-07.
| StepFun: Step 3.7 Flash | Google: Gemma 4 31B | |
|---|---|---|
| Vendor | stepfun | |
| Quality Score | 100 | 100 |
| Benchmark Score | 50.8 | 53.7 |
| Input Price | $0.20/M | $0.20/M |
| Output Price | $1.15/M | $0.40/M |
| Context Window | 262,144 | 262,144 |
| Max Output | 230,400 | 16,384 |
| Tool Calling | ✓ | ✓ |
| Structured Output | ✓ | ✓ |
| Reasoning Mode | ✓ | ✓ |
| Vision | ✓ | ✓ |
| Audio | - | - |
| Benchmark Scores | ||
| ai_index | 49.9 | 48.4 |
| ai_index_agentic | 35.5 | 23.8 |
| ai_index_coding | 65.3 | 71.7 |
| eqbench | - | 57.5 |
The verdict
For a sample job of 1 million input tokens and 250,000 output tokens, Gemma 4 31B costs $0.30 and Step 3.7 Flash costs $0.49, so Gemma 4 31B is about 1.6 times cheaper for the same work.
Both read up to 262,144 tokens at once, roughly 393 pages of text.
On our blended benchmark score Gemma 4 31B leads 53.7 to 50.8.
Across the 20 tasks where they differ, Gemma 4 31B on 20 (SQL Generation, Code Review, Code Completion and others).
Step 3.7 Flash was released in May 2026 and Gemma 4 31B in April 2026.
Pick Gemma 4 31B if you need lower cost, benchmark performance or the stronger all-round task scores.
Costs use current list prices per million tokens; page counts assume about 0.75 words per token and 500 words per page.
Who wins by task?
| Task | StepFun: Step 3.7 Flash | Google: Gemma 4 31B |
|---|---|---|
| SQL Generation | 153 | 156 |
| Code Review | 146 | 152 |
| Code Completion | 130 | 131 |
| Code Refactoring | 144 | 151 |
| Bug Fixing | 155 | 160 |
| Unit Test Generation | 139 | 143 |
| Code Documentation | 133 | 137 |
| Regex Writing | 129 | 131 |
| CI/CD Pipelines | 131 | 135 |
| Frontend Component Design | 136 | 138 |
| Data Analysis | 150 | 152 |
| CSV / Spreadsheet Cleanup | 141 | 145 |
| ETL Scripting | 137 | 143 |
| JSON Extraction | 142 | 143 |
| Bulk Data Labeling | 133 | 133 |
| OCR / Document Parsing | 138 | 140 |
| Table Extraction from PDFs | 138 | 140 |
| Long-Document Summarization | 142 | 149 |
| Short-Form Summarization | 128 | 129 |
| Blog Post Writing | 129 | 133 |
Scores reflect capability match + benchmark data + pricing for each task. Methodology →