Head-to-head comparison
Qwen3 30B A3B Instruct 2507 (Qwen) and Step 3.7 Flash (StepFun) compared on the ModelCap Index, API price, context window, provider availability, weight access and every public benchmark board they share. Figures are the same ones shown on the live rankings; nothing here is a hidden score.
Snapshot as of 15 September 2026
As of 15 September 2026, Step 3.7 Flash holds the stronger ModelCap Index position (#97 vs #100); and Qwen3 30B A3B Instruct 2507 is 6× cheaper per output token ($0.193/1M vs $1.15/1M).
Evidence: 2 public benchmark observations across 2 boards · 3 reported rows against measured corpus ladders (11%); global-corpus-prior over 194 held-out anchors (20%); measured predecessor stepfun/step-3.5-flash less the succession penalty (69%); shrunk 1.1 up toward the measured corpus. Prices are the lowest listed API offer per million tokens observed on the OpenRouter catalogue.
Scores are the sources' own published figures for each model's best evaluated configuration; ModelCap never re-runs a benchmark.
Want a different pairing? Open the interactive comparison tool to swap either model for any current ranked language model.
Step 3.7 Flash ranks higher on the ModelCap Index as of 15 September 2026: #97 against #100. That is a capability ranking built from public benchmark evidence with published uncertainty; whether it is "better" for you also depends on price, context and where you can run it. Their published uncertainty intervals overlap, so the rank difference alone does not establish a reliable capability advantage for your workload.
Qwen3 30B A3B Instruct 2507 is cheaper on output tokens: $0.193/1M against $1.15/1M. Input tokens are $0.048/1M for Qwen3 30B A3B Instruct 2507 and $0.20/1M for Step 3.7 Flash. Prices are the lowest listed API offer ModelCap observed, in USD per million tokens. For 1,000 requests with 2,000 input and 500 output tokens each (2M input + 0.5M output), the listed-rate estimate is $0.19 for Qwen3 30B A3B Instruct 2507 versus $0.97 for Step 3.7 Flash. Qwen3 30B A3B Instruct 2507 costs 80.2% less in this scenario. This excludes caching, batch discounts, prompt-length tiers, tool charges and retries; verify the selected endpoint before budgeting.
Both publish a 262K-token context window.
The two models do not share a coding benchmark board on ModelCap yet, so no head-to-head coding score is published; the ModelCap Index position is the closest overall signal.
Qwen3 30B A3B Instruct 2507: Open weights. Step 3.7 Flash: Open weights. Open weights mean the checkpoint can be downloaded and self-hosted under its licence; API-only models are available solely through hosted endpoints.
ModelCap currently lists 5 API providers for Qwen3 30B A3B Instruct 2507 and 3 for Step 3.7 Flash, from the OpenRouter catalogue snapshot the site serves; each model page lists the providers and their prices.
Every figure comes from the sealed ModelCap dataset published 15 September 2026; the page re-renders within a minute of each data refresh, and the ModelCap Index positions are the same ones shown on the live rankings.