DeepSeek V3.2 vs Qwen3.7 Flash vs Qwen3.8 2.4T A95B
No independent capability benchmark covers every model here yet. The score uses price, inputs & features and context window.
Price in rating: Included · Exclude price
DeepSeek
DeepSeek V3.2
52/100- ECI146.3
- Price$0.28 / $0.42
- Context128K
Alibaba (Qwen)
Qwen3.7 Flash
87/100- ECI—
- Price$0.03 / $0.13
- Context1M
Alibaba (Qwen)
Qwen3.8 2.4T A95B
32/100- ECI—
- Price$2.00 / $6.00
- Context262K
Capability evidence is incomplete
There is not enough shared capability evidence to name an overall winner. The scores (Qwen3.7 Flash 87/100, DeepSeek V3.2 52/100, Qwen3.8 2.4T A95B 32/100) compare the available specifications and 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 priceQwen3.7 FlashQwen3.7 Flash $0.055 · DeepSeek V3.2 $0.315 · Qwen3.8 2.4T A95B $3.00 per 1M tokens (3:1 blend)
- Longest contextQwen3.7 FlashQwen3.7 Flash 1,000,000 · Qwen3.8 2.4T A95B 262,144 · DeepSeek V3.2 128,000 tokens
- Widest inputsQwen3.7 FlashDeepSeek V3.2: Text · Qwen3.7 Flash: Text, Images, Video · Qwen3.8 2.4T A95B: Text
- Self-hostingDeepSeek V3.2 and Qwen3.8 2.4T A95BPublishes downloadable weights (MIT License and qwen3.8-max)
| Measure | Weight | DeepSeek V3.2 | Qwen3.7 Flash | Qwen3.8 2.4T A95B |
|---|---|---|---|---|
| Price | 50% | 71 | 100 | 26 |
| Inputs & features | 30% | 35 | 80 | 35 |
| Context window | 20% | 32 | 64 | 43 |
| Specs & price only | 100% | 52/100 | 87/100 | 32/100 |
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 | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 146.3 | — | — |
| ECI rank | #87 of 158 | — | — |
| GPQA DiamondGraduate-level science questions | 83.4% (best) | 82.3% | — |
| FrontierMath Tiers 1–3Research-level mathematics | — | 19.3% | — |
| OTIS Mock AIME 2024–2025Competition mathematics | 87.8% (best) | 86.7% | — |
| Price per million tokens | |||
| Input | $0.28 | $0.03 (best) | $2.00 |
| Output | $0.42 | $0.13 (best) | $6.00 |
| Cached input | — | $0.006 | — |
| Blended (3:1) | $0.315 | $0.055 (best) | $3.00 |
| Long-context rate | Same rate | Over 32K: $0.10 / $0.40 | Same rate |
| Price source | Median of 30 providers | Official Alibaba API | Median of 26 providers |
| Limits | |||
| Context window | 128,000 tokens | 1,000,000 tokens (best) | 262,144 tokens |
| Max output | 64,000 tokens | 131,072 tokens (best) | 131,072 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | Yes | No |
| Reasoning | Yes | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | No |
| Availability | |||
| Weights | OpenMIT License | Proprietary | Openqwen3.8-max |
| API model ID | — | qwen3.7-flash | — |
| API providers | 30 (best) | 15 | 26 |
| Released | Dec 1, 2025 | Jul 15, 2026 | Aug 12, 2026 |
| Knowledge cutoff | Jul 2024 | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
DeepSeek V3.2$3.64
Qwen3.7 Flash$0.56
Qwen3.8 2.4T A95B$32.00
Which should you choose?
Which is better: DeepSeek V3.2, Qwen3.7 Flash or Qwen3.8 2.4T A95B?
There is not enough shared capability evidence to name an overall winner. The scores (Qwen3.7 Flash 87/100, DeepSeek V3.2 52/100, Qwen3.8 2.4T A95B 32/100) compare the available specifications and 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, DeepSeek V3.2, Qwen3.7 Flash or Qwen3.8 2.4T A95B?
Qwen3.7 Flash is cheaper at $0.03 input / $0.13 output per million tokens (official Alibaba API price). DeepSeek V3.2 costs $0.28 input / $0.42 output per million tokens (median across 30 API providers); Qwen3.8 2.4T A95B costs $2.00 input / $6.00 output per million tokens (median across 26 API providers). At a typical mix of three input tokens to one output token, that is $0.055 per million tokens for Qwen3.7 Flash versus $0.315 for DeepSeek V3.2 (5.7× as much) and $3.00 for Qwen3.8 2.4T A95B (55× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. DeepSeek V3.2 has an ECI of 146.3, Qwen3.7 Flash has not been scored yet and Qwen3.8 2.4T A95B has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for DeepSeek V3.2, Qwen3.7 Flash and Qwen3.8 2.4T A95B 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?
Qwen3.7 Flash has the largest context window at 1,000,000 tokens, against 262,144 for Qwen3.8 2.4T A95B and 128,000 for DeepSeek V3.2. Maximum output per response: DeepSeek V3.2 up to 64,000, Qwen3.7 Flash up to 131,072, Qwen3.8 2.4T A95B up to 131,072 tokens.
Which can read images, PDFs, audio or video?
DeepSeek V3.2 accepts text; Qwen3.7 Flash accepts text, images and video; Qwen3.8 2.4T A95B accepts text. Qwen3.7 Flash handles the widest range of inputs.
Are any of these open source?
DeepSeek V3.2 and Qwen3.8 2.4T A95B publishes its weights (MIT License and qwen3.8-max) and can be self-hosted; Qwen3.7 Flash is proprietary.
Which is newer?
Qwen3.8 2.4T A95B is the newest, released Aug 12, 2026. Qwen3.7 Flash came out Jul 15, 2026; DeepSeek V3.2 came out Dec 1, 2025. Knowledge cutoff: DeepSeek V3.2 Jul 2024.
How do you decide the winner?
This rating uses price (50%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (30%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (20%, 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.