Qwen3.5-Omni-Flash leads on the shared-benchmark average
Ahead on 2 of 2 shared benchmarks, averaging 30.6 points higher
Summarised only from the 2 percentage-scale benchmarks scored by every selected model; details are below.
“Best available” takes each model’s best recorded non-parallel result in the metric’s stated direction per benchmark, so it may combine modes into a virtual configuration that does not exist. Read it with the mode breakdown.

Gemma 4 E4B
DeepMind
Benchmark-by-benchmark comparison. Changing the thinking mode or tool filters updates the chart and table below.
“Best available” picks the best non-parallel result in the metric’s stated direction separately for each benchmark. The resulting series can combine several reasoning levels and is not one reproducible runtime configuration. Choose a mode filter to compare like-for-like runs.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Each axis is the mean percentage score of one benchmark domain. It is an average, not a capability rating.
Relative edge: 指令跟随 +4.4 / Relative gap: Agent能力评测 -30.7
Relative edge: Agent能力评测 +30.7 / Relative gap: 指令跟随 -4.4
Method: for each model and benchmark, all scores in the current mode scope are averaged (not the best score), then those benchmark scores are averaged within each domain. Only benchmarks scored on a 0-100 scale by at least two of the selected models count — Elo and rating-scale benchmarks such as Codeforces or Arena are excluded, because averaging a 1500 rating with an 85% accuracy produces a meaningless number. Missing values are not counted as zero, and the averages are unweighted, so domains with harder benchmarks read lower.
Every model and runtime mode, benchmark by benchmark. Values are comparable along a row, not between different benchmarks.
6 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | Gemma 4 E4B | Qwen3.5-Omni-Flash |
|---|---|---|
4.80Standard Mode | 7.60Standard Mode | |
58.60Thinking Enabled | 74.20Standard Mode | |
8.30Thinking Enabled | Tools | 8.30Standard Mode | Tools | |
26.00Standard Mode | Tools | 84.50Standard Mode | Tools | |
44.20Thinking Enabled | 38.00Standard Mode | |
52.60Thinking Level · High | 64.70Standard Mode |
Official list prices per model API, split by input and output. Unit: USD per 1M tokens.
Architecture, licensing and API modalities. "Not provided" means the field is missing from our database.
| Features & specs | Gemma 4 E4BDeepMind | Qwen3.5-Omni-Flash阿里巴巴 |
|---|---|---|
Core specsRelease | 2026-04-02 | 2026-03-30 |
Context length | 128K | 256K |
Total parameters | 8B | — |
Max output length | 8,192 tokens | 8,192 tokens |
Architecture | Dense | Undisclosed |
Runtime modes | 关闭开启 | 关闭开启 |
Availability & licensingCode availability | Available · Apache 2.0 | Available · Qwen License |
Weight availability | Available · Apache 2.0 | Available · Qwen License |
Use & commercial terms | 免费商用授权 | 免费商用授权 |
Local deploymentWeights | Hugging Face | Not provided |
API modality supportText Input/Output | Input:YesOutput:Yes | Input:YesOutput:Yes |
Image Input/Output | Input:YesOutput:No | Input:YesOutput:No |
Audio Input/Output | Input:YesOutput:No | Input:YesOutput:Yes |
Video Input/Output | Input:NoOutput:No | Input:YesOutput:No |
ResourcesPaper / report | Gemma 4 Model Card | Qwen3.5-Omni: Scaling Up, Toward Native Omni-Modal AGI |
DataLearner blog | Google Gemma 4 正式开源:Apache 2.0 协议、手机端可运行、原生支持多模态和 Agent 工作流 | Not provided |

Qwen3.5-Omni-Flash
阿里巴巴