Step 3.7 Flash is 22.2× cheaper per million tokens (blended). Claude Opus 5 has the larger context window (1,000,000 vs 256,000 tokens, 3.9×). Step 3.7 Flash is open-weights; the other is not.
Claude Opus 5 vs Step 3.7 Flash
Live catalog fields. Best-in-row is highlighted.
Claude Opus 5
Anthropic · anthropic__claude-opus-5
Step 3.7 Flashopen
Stepfun · stepfun__step-3.7-flash
Identity
| Field | anthropic__claude-opus-5 | stepfun__step-3.7-flash |
|---|---|---|
| Provider | Anthropic | Stepfun |
| Slug | anthropic__claude-opus-5 | stepfun__step-3.7-flash |
| Status | live | live |
| Open weights | No | Yes |
| License | — | — |
| HuggingFace | — | stepfun-ai/Step-3.7-Flash |
Quality
| Field | anthropic__claude-opus-5 | stepfun__step-3.7-flash |
|---|---|---|
| Quality index (0–100) | — | — |
Cost
| Field | anthropic__claude-opus-5 | stepfun__step-3.7-flash |
|---|---|---|
| Input $ / 1M tokens | $5.00 | $0.20 |
| Cached input $ / 1M | $0.50 | $0.04 |
| Output $ / 1M tokens | $25.00 | $1.15 |
Context
| Field | anthropic__claude-opus-5 | stepfun__step-3.7-flash |
|---|---|---|
| Max context tokens | 1,000,000 | 256,000 |
| Max output tokens | 128,000 | 230,400 |
Modalities
| Field | anthropic__claude-opus-5 | stepfun__step-3.7-flash |
|---|---|---|
| Input | text, image, file | text, image, video |
| Output | text | text |