StepFun: Step 5 Preview vs Anthropic: Claude Sonnet 5.5 (batch)
Side-by-side comparison of specs, pricing, benchmark scores, and task rankings. Updated 2026-11-11.
| StepFun: Step 5 Preview | Anthropic: Claude Sonnet 5.5 (batch) | |
|---|---|---|
| Vendor | stepfun | anthropic |
| Quality Score | 100 | 100 |
| Benchmark Score | 43.5 | 89.3 |
| Input Price | $1.00/M | $1.00/M |
| Output Price | $2.70/M | $5.00/M |
| Context Window | 1,000,000 | 1,000,000 |
| Max Output | 64,000 | 128,000 |
| Tool Calling | ✓ | ✓ |
| Structured Output | ✓ | ✓ |
| Reasoning Mode | ✓ | ✓ |
| Vision | ✓ | ✓ |
| Audio | - | - |
| Benchmark Scores | ||
| ai_index | 47.1 | 90.0 |
The verdict
For a sample job of 1 million input tokens and 250,000 output tokens, Step 5 Preview costs $1.68 and Claude Sonnet 5.5 (batch) costs $2.25, so Step 5 Preview is about 1.3 times cheaper for the same work.
Both read up to 1,000,000 tokens at once, roughly 1,500 pages of text.
Only Step 5 Preview offers video input.
On our blended benchmark score Claude Sonnet 5.5 (batch) leads 89.3 to 43.5.
Across the 20 tasks where they differ, Step 5 Preview scores higher on 2 (JSON Extraction, Bulk Data Labeling), Claude Sonnet 5.5 (batch) on 18 (SQL Generation, Code Review, Code Completion and others).
Step 5 Preview was released in October 2026 and Claude Sonnet 5.5 (batch) in September 2026.
Pick Step 5 Preview if you need lower cost or video input; pick Claude Sonnet 5.5 (batch) if you need 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 5 Preview | Anthropic: Claude Sonnet 5.5 (batch) |
|---|---|---|
| SQL Generation | 139 | 144 |
| Code Review | 140 | 149 |
| Code Completion | 120 | 120 |
| Code Refactoring | 143 | 150 |
| Bug Fixing | 144 | 153 |
| Unit Test Generation | 129 | 134 |
| Code Documentation | 135 | 138 |
| Regex Writing | 124 | 127 |
| CI/CD Pipelines | 125 | 130 |
| Frontend Component Design | 127 | 132 |
| Data Analysis | 131 | 137 |
| CSV / Spreadsheet Cleanup | 135 | 137 |
| ETL Scripting | 134 | 140 |
| JSON Extraction | 123 | 122 |
| Bulk Data Labeling | 120 | 120 |
| OCR / Document Parsing | 133 | 135 |
| Table Extraction from PDFs | 133 | 135 |
| Long-Document Summarization | 145 | 151 |
| Short-Form Summarization | 117 | 119 |
| Blog Post Writing | 127 | 132 |
Scores reflect capability match + benchmark data + pricing for each task. Methodology →