Hy4 flash vs MAI Code 2 Flash vs Step 5 Preview
No independent capability benchmark covers every model here yet. The score uses context window.
Price in rating: Excluded · Include price
Tencent
Hy4 flash
65/100- ECI—
- Price$0.07 / $0.281
- Context1.05M
Microsoft
MAI Code 2 Flash
65/100- ECI—
- Price$0.05 / $0.25
- Context1.05M
StepFun
Step 5 Preview
64/100- ECI—
- Price$1.00 / $2.70
- Context1M
Capability evidence is incomplete
There is not enough shared capability evidence to name an overall winner. The scores (MAI Code 2 Flash 65/100, Hy4 flash 65/100, Step 5 Preview 64/100) compare the available specifications only. 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 priceMAI Code 2 FlashMAI Code 2 Flash $0.10 · Hy4 flash $0.123 · Step 5 Preview $1.43 per 1M tokens (3:1 blend)
- Longest contextMAI Code 2 Flash and Hy4 flashMAI Code 2 Flash 1,050,000 · Hy4 flash 1,048,576 · Step 5 Preview 1,000,000 tokens
- Widest inputsStep 5 PreviewHy4 flash: Text, Images · MAI Code 2 Flash: Text, Images · Step 5 Preview: Text, Images, Video
- Self-hostingStep 5 PreviewPublishes downloadable weights
| Measure | Weight | Hy4 flash | MAI Code 2 Flash | Step 5 Preview |
|---|---|---|---|---|
| Context window | 100% | 65 | 65 | 64 |
| Specs only | 100% | 65/100 | 65/100 | 64/100 |
Left out because at least one model lacks the data: capability 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 | — | Coding | — |
| Published capabilities | Tencent's lower-cost Hy4 tier for everyday agent and productivity tasks, with image understanding and a 1M-token context window. | Smaller mixture-of-experts coding model for quick edits, explanations and iteration. | 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 |
| Model size | — | 12B active parameters | — |
| Access | — | GitHub Copilot and limited-access Microsoft Foundry preview | — |
| Billing terms | — | Promotional GitHub Copilot token rates until May 17, 2027. | — |
| Sources | — | Publisher documentation · Additional documentation · Pricing documentation | — |
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.07 | $0.05 (best) | $1.00 |
| Output | $0.281 | $0.25 (best) | $2.70 |
| Cached input | — | $0.005 (best) | $0.10 |
| Blended (3:1) | $0.123 | $0.10 (best) | $1.43 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 10 providers | Official GitHub Copilot API | Official StepFun (Global) API |
| Limits | |||
| Context window | 1,048,576 tokens | 1,050,000 tokens (best) | 1,000,000 tokens |
| Max output | 65,536 tokens | 128,000 tokens | 1,000,000 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | Yes | Yes | Yeslow · medium · high |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | — | Yes |
| Availability | |||
| Weights | Proprietary | Proprietary | Open |
| API model ID | — | — | step-5-preview |
| API providers | 11 | — | 21 (best) |
| Released | Nov 16, 2026 | Nov 17, 2026 | Sep 16, 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.
Hy4 flash$1.26
MAI Code 2 Flash$1.00
Step 5 Preview$15.40
Which should you choose?
Which is better: Hy4 flash, MAI Code 2 Flash or Step 5 Preview?
There is not enough shared capability evidence to name an overall winner. The scores (MAI Code 2 Flash 65/100, Hy4 flash 65/100, Step 5 Preview 64/100) compare the available specifications only. 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, Hy4 flash, MAI Code 2 Flash or Step 5 Preview?
MAI Code 2 Flash is cheaper at $0.05 input / $0.25 output per million tokens (official GitHub Copilot API price). Hy4 flash costs $0.07 input / $0.281 output per million tokens (median across 10 API providers); 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 $0.10 per million tokens for MAI Code 2 Flash versus $0.123 for Hy4 flash (1.2× as much) and $1.43 for Step 5 Preview (14× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Hy4 flash has not been scored yet, MAI Code 2 Flash has not been scored yet and Step 5 Preview has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Hy4 flash, MAI Code 2 Flash and Step 5 Preview yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
MAI Code 2 Flash and Hy4 flash have the largest context windows (1,050,000 and 1,048,576 tokens), against 1,000,000 for Step 5 Preview. Maximum output per response: Hy4 flash up to 65,536, MAI Code 2 Flash up to 128,000, Step 5 Preview up to 1,000,000 tokens.
Which can read images, PDFs, audio or video?
Hy4 flash accepts text and images; MAI Code 2 Flash accepts text and images; Step 5 Preview accepts text, images and video. Step 5 Preview handles the widest range of inputs.
Are any of these open source?
Step 5 Preview publishes its weights and can be self-hosted; Hy4 flash and MAI Code 2 Flash is proprietary.
Which is newer?
MAI Code 2 Flash is the newest, released Nov 17, 2026. Hy4 flash came out Nov 16, 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.