MiniCPM5-2B vs Qwen3.8 Flash Next vs Step 3.7 Flash
Qwen3.8 Flash Next comes out ahead, 70 to 62 and 53 on our weighted score, and it is the cheaper option too.
OpenBMB
MiniCPM5-2B
53/100- ECI—
- Price$0.124 / $0.743
- Context131K
- Our pick
Alibaba (Qwen)Qwen3.8 Flash Next
70/100- ECI—
- Price$0.20 / $0.50
- Context262K
StepFunStep 3.7 Flash
62/100- ECI—
- Price$0.185 / $1.11
- Context256K
Qwen3.8 Flash Next is our pick
Qwen3.8 Flash Next is the better all-round choice, scoring 70/100 against Step 3.7 Flash (62) and MiniCPM5-2B (53). It leads on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. 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 priceQwen3.8 Flash NextQwen3.8 Flash Next $0.275 · MiniCPM5-2B $0.279 · Step 3.7 Flash $0.416 per 1M tokens (3:1 blend)
- Longest contextQwen3.8 Flash NextQwen3.8 Flash Next 262,144 · Step 3.7 Flash 256,000 · MiniCPM5-2B 131,072 tokens
- Widest inputsQwen3.8 Flash Next and Step 3.7 FlashMiniCPM5-2B: Text · Qwen3.8 Flash Next: Text, Images, Video · Step 3.7 Flash: Text, Images, Video
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | MiniCPM5-2B | Qwen3.8 Flash Next | Step 3.7 Flash |
|---|---|---|---|---|
| Price | 50% | 76 | 76 | 68 |
| Inputs & features | 30% | 35 | 80 | 70 |
| Context window | 20% | 24 | 37 | 36 |
| Overall | 100% | 53/100 | 70/100 | 62/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | MiniCPM5-2BOpenBMB | Qwen3.8 Flash NextAlibaba (Qwen) | Step 3.7 FlashStepFun |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.124 (best) | $0.20 | $0.185 |
| Output | $0.743 | $0.50 (best) | $1.11 |
| Cached input | — | — | $0.037 |
| Blended (3:1) | $0.279 | $0.275 (best) | $0.416 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 1 providers | Median of 5 providers | Official StepFun (Global) API |
| Limits | |||
| Context window | 131,072 tokens | 262,144 tokens (best) | 256,000 tokens |
| Max output | 131,072 tokens | 131,072 tokens | 256,000 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | Yes | Yes |
| Reasoning | Yes | Yes | Yeslow · medium · high |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | No |
| Availability | |||
| Weights | Openapache-2.0 | Openqwen-community-1.0 | Open |
| API model ID | — | — | step-3.7-flash |
| API providers | 1 | 5 | 18 (best) |
| Released | Sep 6, 2026 | Aug 27, 2026 | May 29, 2026 |
| Knowledge cutoff | — | — | Mar 1, 2026 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- MiniCPM5-2B$2.73
- Qwen3.8 Flash Next$3.00
- Step 3.7 Flash$4.07
Which should you choose?
Which is better: MiniCPM5-2B, Qwen3.8 Flash Next or Step 3.7 Flash?
Qwen3.8 Flash Next is the better all-round choice, scoring 70/100 against Step 3.7 Flash (62) and MiniCPM5-2B (53). It leads on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, MiniCPM5-2B, Qwen3.8 Flash Next or Step 3.7 Flash?
Qwen3.8 Flash Next is cheaper at $0.20 input / $0.50 output per million tokens (median across 5 API providers). MiniCPM5-2B costs $0.124 input / $0.743 output per million tokens (median across 1 API provider); Step 3.7 Flash costs $0.185 input / $1.11 output per million tokens (official StepFun (Global) API price). At a typical mix of three input tokens to one output token, that is $0.275 per million tokens for Qwen3.8 Flash Next versus $0.279 for MiniCPM5-2B (1× as much) and $0.416 for Step 3.7 Flash (1.5× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. MiniCPM5-2B has not been scored yet, Qwen3.8 Flash Next has not been scored yet and Step 3.7 Flash has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for MiniCPM5-2B, Qwen3.8 Flash Next and Step 3.7 Flash 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?
Qwen3.8 Flash Next has the largest context window at 262,144 tokens, against 256,000 for Step 3.7 Flash and 131,072 for MiniCPM5-2B. Maximum output per response: MiniCPM5-2B up to 131,072, Qwen3.8 Flash Next up to 131,072, Step 3.7 Flash up to 256,000 tokens.
Which can read images, PDFs, audio or video?
MiniCPM5-2B accepts text; Qwen3.8 Flash Next accepts text, images and video; Step 3.7 Flash accepts text, images and video. Qwen3.8 Flash Next handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights (apache-2.0 and qwen-community-1.0), so you can self-host them.
Which is newer?
MiniCPM5-2B is the newest, released Sep 6, 2026. Qwen3.8 Flash Next came out Aug 27, 2026; Step 3.7 Flash came out May 29, 2026. Knowledge cutoff: Step 3.7 Flash Mar 1, 2026.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, 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 Oct 4, 2026.
