DeepSeek V4 Flash vs Gemma 4 31B IT vs Qwen3.5 9B
Too close to call on our weighted score (Gemma 4 31B IT 68, DeepSeek V4 Flash 65, Qwen3.5 9B 65). The right pick depends on what you value most.
Price in rating: Excluded · Include price
DeepSeek
DeepSeek V4 Flash
65/100- ECI146.8
- Price$0.14 / $0.28
- Context1M
Google
Gemma 4 31B IT
68/100- ECI144.1
- Price$0.142 / $0.40
- Context262K
Alibaba (Qwen)
Qwen3.5 9B
65/100- ECI139.5
- Price$0.10 / $0.19
- Context262K
Too close to call
It is close. Our weighted score puts them within 2 points (Gemma 4 31B IT 68/100, DeepSeek V4 Flash 65/100, Qwen3.5 9B 65/100), so choose by what matters most for your work: DeepSeek V4 Flash for raw capability and Qwen3.5 9B on price. The score weighs capability 67%, inputs & features 20%, context window 13%.
- CapabilityDeepSeek V4 FlashCapabilities Index (ECI): DeepSeek V4 Flash 146.8 · Gemma 4 31B IT 144.1 · Qwen3.5 9B 139.5
- Lowest priceQwen3.5 9BQwen3.5 9B $0.122 · DeepSeek V4 Flash $0.175 · Gemma 4 31B IT $0.206 per 1M tokens (3:1 blend)
- Longest contextDeepSeek V4 FlashDeepSeek V4 Flash 1,000,000 · Gemma 4 31B IT 262,144 · Qwen3.5 9B 262,144 tokens
- Widest inputsQwen3.5 9BDeepSeek V4 Flash: Text · Gemma 4 31B IT: Text, Images · Qwen3.5 9B: Text, Images, Video
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | DeepSeek V4 Flash | Gemma 4 31B IT | Qwen3.5 9B |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 67% | 74 | 72 | 67 |
| Inputs & features | 20% | 35 | 70 | 70 |
| Context window | 13% | 64 | 43 | 43 |
| Overall | 100% | 65/100 | 68/100 | 65/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 146.8 (best) | 144.1 | 139.5 |
| ECI rank | #83 of 158 (best) | #94 of 158 | #116 of 158 |
| GPQA DiamondGraduate-level science questions | — | 75.8% | 79.0% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | — | 73.3% (best) | 61.7% |
| SimpleQA VerifiedShort factual questions | — | 10.4% | — |
| Price per million tokens | |||
| Input | $0.14 | $0.142 | $0.10 (best) |
| Output | $0.28 | $0.40 | $0.19 (best) |
| Cached input | — | — | — |
| Blended (3:1) | $0.175 | $0.206 | $0.122 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 52 providers | Median of 42 providers | Median of 20 providers |
| Limits | |||
| Context window | 1,000,000 tokens (best) | 262,144 tokens | 262,144 tokens |
| Max output | 384,000 tokens (best) | 32,768 tokens | 65,536 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | Yes | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | No |
| Availability | |||
| Weights | Open | Open | Open |
| API model ID | — | gemma-4-31b-it | — |
| API providers | 58 (best) | 52 | 22 |
| Released | Apr 24, 2026 | Apr 2, 2026 | Feb 23, 2026 |
| Knowledge cutoff | May 2025 | 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.
DeepSeek V4 Flash$1.96
Gemma 4 31B IT$2.22
Qwen3.5 9B$1.38
Which should you choose?
Which is better: DeepSeek V4 Flash, Gemma 4 31B IT or Qwen3.5 9B?
It is close. Our weighted score puts them within 2 points (Gemma 4 31B IT 68/100, DeepSeek V4 Flash 65/100, Qwen3.5 9B 65/100), so choose by what matters most for your work: DeepSeek V4 Flash for raw capability and Qwen3.5 9B on price. The score weighs capability 67%, inputs & features 20%, context window 13%.
Which is cheaper, DeepSeek V4 Flash, Gemma 4 31B IT or Qwen3.5 9B?
Qwen3.5 9B is cheaper at $0.10 input / $0.19 output per million tokens (median across 20 API providers). DeepSeek V4 Flash costs $0.14 input / $0.28 output per million tokens (median across 52 API providers); Gemma 4 31B IT costs $0.142 input / $0.40 output per million tokens (median across 42 API providers; free on Google). At a typical mix of three input tokens to one output token, that is $0.122 per million tokens for Qwen3.5 9B versus $0.175 for DeepSeek V4 Flash (1.4× as much) and $0.206 for Gemma 4 31B IT (1.7× as much).
Which scores higher on benchmarks?
DeepSeek V4 Flash scores higher on the Capabilities Index (ECI): DeepSeek V4 Flash 146.8 (#83 of 158), Gemma 4 31B IT 144.1 (#94 of 158) and Qwen3.5 9B 139.5 (#116 of 158). The confidence ranges of the top two overlap (143.6–150.2 vs 141.7–146.6), so treat the gap as small.
Which is better for coding?
There are no published SWE-bench Verified results for DeepSeek V4 Flash, Gemma 4 31B IT and Qwen3.5 9B yet, so there is no like-for-like coding score. On overall capability, DeepSeek V4 Flash leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.
Which has the bigger context window?
DeepSeek V4 Flash has the largest context window at 1,000,000 tokens, against 262,144 for Gemma 4 31B IT and 262,144 for Qwen3.5 9B. Maximum output per response: DeepSeek V4 Flash up to 384,000, Gemma 4 31B IT up to 32,768, Qwen3.5 9B up to 65,536 tokens.
Which can read images, PDFs, audio or video?
DeepSeek V4 Flash accepts text; Gemma 4 31B IT accepts text and images; Qwen3.5 9B accepts text, images and video. Qwen3.5 9B handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights, so you can self-host them.
Which is newer?
DeepSeek V4 Flash is the newest, released Apr 24, 2026. Gemma 4 31B IT came out Apr 2, 2026; Qwen3.5 9B came out Feb 23, 2026. Knowledge cutoff: DeepSeek V4 Flash May 2025, Gemma 4 31B IT Jan 2025.
How do you decide the winner?
This rating uses capability (67%, independent benchmark results); inputs & features (20%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (13%, 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 18, 2026.