Qwen3.5 Flash vs Step 3.5 Flash 2603 vs Gemma 4 31B IT
Qwen3.5 Flash comes out ahead, 79 to 70 and 62 on our weighted score, though Step 3.5 Flash 2603 is 14% cheaper per token.
- Our pick
Alibaba (Qwen)Qwen3.5 Flash
79/100- ECI144.0
- Price$0.10 / $0.40
- Context1M
StepFunStep 3.5 Flash 2603
62/100- ECI—
- Price$0.10 / $0.30
- Context256K
GoogleGemma 4 31B IT
70/100- ECI142.8
- Price$0.14 / $0.40
- Context262K
Qwen3.5 Flash is our pick
Qwen3.5 Flash is the better all-round choice, scoring 79/100 against Gemma 4 31B IT (70) and Step 3.5 Flash 2603 (62). It leads on inputs & features and context window. Step 3.5 Flash 2603 wins on price. 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 priceStep 3.5 Flash 2603Step 3.5 Flash 2603 $0.15 · Qwen3.5 Flash $0.175 · Gemma 4 31B IT $0.205 per 1M tokens (3:1 blend)
- Longest contextQwen3.5 FlashQwen3.5 Flash 1,000,000 · Gemma 4 31B IT 262,144 · Step 3.5 Flash 2603 256,000 tokens
- Widest inputsQwen3.5 FlashQwen3.5 Flash: Text, Images, Video · Step 3.5 Flash 2603: Text · Gemma 4 31B IT: Text, Images
- Self-hostingStep 3.5 Flash 2603 and Gemma 4 31B ITPublishes downloadable weights
| Measure | Weight | Qwen3.5 Flash | Step 3.5 Flash 2603 | Gemma 4 31B IT |
|---|---|---|---|---|
| Price | 50% | 86 | 89 | 83 |
| Inputs & features | 30% | 80 | 35 | 70 |
| Context window | 20% | 60 | 36 | 37 |
| Overall | 100% | 79/100 | 62/100 | 70/100 |
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 | Qwen3.5 FlashAlibaba (Qwen) | Step 3.5 Flash 2603StepFun | Gemma 4 31B ITGoogle |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 144.0 (best) | — | 142.8 |
| ECI rank | #82 of 148 (best) | — | #86 of 148 |
| GPQA DiamondGraduate-level science questions | 82.3% (best) | — | 75.8% |
| FrontierMath Tiers 1–3Research-level mathematics | 18.3% | — | — |
| OTIS Mock AIME 2024–2025Competition mathematics | 84.4% (best) | — | 73.3% |
| SimpleQA VerifiedShort factual questions | 20.3% (best) | — | 10.4% |
| Price per million tokens | |||
| Input | $0.10 (best) | $0.10 (best) | $0.14 |
| Output | $0.40 | $0.30 (best) | $0.40 |
| Cached input | $0.01 (best) | $0.02 | — |
| Blended (3:1) | $0.175 | $0.15 (best) | $0.205 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Alibaba API | Official StepFun (Global) API | Median of 30 providers |
| Limits | |||
| Context window | 1,000,000 tokens (best) | 256,000 tokens | 262,144 tokens |
| Max output | 65,536 tokens | 256,000 tokens (best) | 32,768 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | No | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | Yes | No | No |
| Reasoning | Yes | Yeslow · high | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | No | Yes |
| Availability | |||
| Weights | Proprietary | Open | Open |
| API model ID | qwen3.5-flash | step-3.5-flash-2603 | gemma-4-31b-it |
| API providers | 8 | 3 | 38 (best) |
| Released | Feb 23, 2026 | Apr 2, 2026 | Apr 2, 2026 |
| Knowledge cutoff | — | Jan 2025 | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Qwen3.5 Flash$1.80
- Step 3.5 Flash 2603$1.60
- Gemma 4 31B IT$2.20
Which should you choose?
Which is better: Qwen3.5 Flash, Step 3.5 Flash 2603 or Gemma 4 31B IT?
Qwen3.5 Flash is the better all-round choice, scoring 79/100 against Gemma 4 31B IT (70) and Step 3.5 Flash 2603 (62). It leads on inputs & features and context window. Step 3.5 Flash 2603 wins on price. 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, Qwen3.5 Flash, Step 3.5 Flash 2603 or Gemma 4 31B IT?
Step 3.5 Flash 2603 is cheaper at $0.10 input / $0.30 output per million tokens (official StepFun (Global) API price). Qwen3.5 Flash costs $0.10 input / $0.40 output per million tokens (official Alibaba API price); Gemma 4 31B IT costs $0.14 input / $0.40 output per million tokens (median across 30 API providers). At a typical mix of three input tokens to one output token, that is $0.15 per million tokens for Step 3.5 Flash 2603 versus $0.175 for Qwen3.5 Flash (1.2× as much) and $0.205 for Gemma 4 31B IT (1.4× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Qwen3.5 Flash has an ECI of 144.0, Step 3.5 Flash 2603 has not been scored yet and Gemma 4 31B IT has an ECI of 142.8.
Which is better for coding?
There are no published SWE-bench Verified results for Qwen3.5 Flash, Step 3.5 Flash 2603 and Gemma 4 31B IT 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?
Qwen3.5 Flash has the largest context window at 1,000,000 tokens, against 262,144 for Gemma 4 31B IT and 256,000 for Step 3.5 Flash 2603. Maximum output per response: Qwen3.5 Flash up to 65,536, Step 3.5 Flash 2603 up to 256,000, Gemma 4 31B IT up to 32,768 tokens.
Which can read images, PDFs, audio or video?
Qwen3.5 Flash accepts text, images and video; Step 3.5 Flash 2603 accepts text; Gemma 4 31B IT accepts text and images. Qwen3.5 Flash handles the widest range of inputs.
Are any of these open source?
Step 3.5 Flash 2603 and Gemma 4 31B IT publishes its weights and can be self-hosted; Qwen3.5 Flash is proprietary.
Which is newer?
Step 3.5 Flash 2603 is the newest, released Apr 2, 2026. Gemma 4 31B IT came out Apr 2, 2026; Qwen3.5 Flash came out Feb 23, 2026. Knowledge cutoff: Step 3.5 Flash 2603 Jan 2025.
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.
