Ripple-65B Preview vs Step 5 Preview vs Hy4
No independent capability benchmark covers every model here yet. The score uses context window.
Price in rating: Included · Exclude price
Ricursive Intelligence
Ripple-65B Preview
65/100- ECI—
- Price—
- Context1.05M
StepFun
Step 5 Preview
64/100- ECI—
- Price$1.00 / $2.70
- Context1M
Tencent
Hy4
65/100- ECI—
- Price$0.834 / $2.50
- Context1.05M
Capability evidence is incomplete
There is not enough shared capability evidence to name an overall winner. The scores (Ripple-65B Preview 65/100, Hy4 65/100, Step 5 Preview 64/100) compare the available specifications and price. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceHy4Hy4 $1.25 · Step 5 Preview $1.43 per 1M tokens (3:1 blend) · Ripple-65B Preview unpriced
- Longest contextRipple-65B Preview and Hy4Ripple-65B Preview 1,048,576 · Hy4 1,048,576 · Step 5 Preview 1,000,000 tokens
- Widest inputsStep 5 PreviewRipple-65B Preview: Text · Step 5 Preview: Text, Images, Video · Hy4: Text, Images
- Self-hostingRipple-65B Preview and Step 5 PreviewPublishes downloadable weights
| Measure | Weight | Ripple-65B Preview | Step 5 Preview | Hy4 |
|---|---|---|---|---|
| Context window | 100% | 65 | 64 | 65 |
| Specs & price only | 100% | 65/100 | 64/100 | 65/100 |
Left out because at least one model lacks the data: capability, price and inputs & features. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Task | Language generation and coding | — | — |
| Published capabilities | Dense transformer trained on 12 trillion tokens with a focus on long-context software engineering. | StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window | A next-generation multimodal productivity model with significantly enhanced Agent and complex task execution capabilities. |
| Model size | 65B parameters | — | — |
| Access | Public weights and hosted inference via Together AI, OpenRouter and Fireworks | — | — |
| Sources | Publisher documentation · Model weights | — | — |
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | — | $1.00 | $0.834 (best) |
| Output | — | $2.70 | $2.50 (best) |
| Cached input | — | $0.10 | $0.042 (best) |
| Blended (3:1) | — | $1.43 | $1.25 (best) |
| Long-context rate | — | Same rate | Same rate |
| Price source | — | Official StepFun (Global) API | Official Tencent TokenHub API |
| Limits | |||
| Context window | 1,048,576 tokens (best) | 1,000,000 tokens | 1,048,576 tokens (best) |
| Max output | — | 1,000,000 tokens (best) | 65,536 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | Yes | No |
| Reasoning | Unconfirmed | Yeslow · medium · high | Yeshigh |
| Tool calling | — | Yes | Yes |
| Structured output | — | Yes | No |
| Availability | |||
| Weights | Open | Open | Proprietary |
| API model ID | — | step-5-preview | hy4 |
| API providers | — | 21 (best) | 12 |
| Released | Nov 4, 2026 | Sep 16, 2026 | Nov 4, 2026 |
| Knowledge cutoff | — | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
Ripple-65B Preview—
Step 5 Preview$15.40
Hy4$13.34
Which should you choose?
Which is better: Ripple-65B Preview, Step 5 Preview or Hy4?
There is not enough shared capability evidence to name an overall winner. The scores (Ripple-65B Preview 65/100, Hy4 65/100, Step 5 Preview 64/100) compare the available specifications and price. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, Ripple-65B Preview, Step 5 Preview or Hy4?
Hy4 is cheaper at $0.834 input / $2.50 output per million tokens (official Tencent TokenHub API price). Step 5 Preview costs $1.00 input / $2.70 output per million tokens (official StepFun (Global) API price). At a typical mix of three input tokens to one output token, that is $1.25 per million tokens for Hy4 versus $1.43 for Step 5 Preview (1.1× as much). Ripple-65B Preview has no published per-token price.
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Ripple-65B Preview has not been scored yet, Step 5 Preview has not been scored yet and Hy4 has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Ripple-65B Preview, Step 5 Preview and Hy4 yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. Note that Ripple-65B Preview has no confirmed tool-calling support in this comparison.
Which has the bigger context window?
Ripple-65B Preview and Hy4 have the largest context windows (1,048,576 and 1,048,576 tokens), against 1,000,000 for Step 5 Preview. Maximum output per response: Step 5 Preview up to 1,000,000, Hy4 up to 65,536 tokens.
Which can read images, PDFs, audio or video?
Ripple-65B Preview accepts text; Step 5 Preview accepts text, images and video; Hy4 accepts text and images. Step 5 Preview handles the widest range of inputs.
Are any of these open source?
Ripple-65B Preview and Step 5 Preview publishes its weights and can be self-hosted; Hy4 is proprietary.
Which is newer?
Ripple-65B Preview is the newest, released Nov 4, 2026. Hy4 came out Nov 4, 2026; Step 5 Preview came out Sep 16, 2026.
How do you decide the winner?
This rating uses context window (100%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 19, 2026.