DeepSeek V3.2 vs GPT-5.4 nano vs Qwen3.8 Max
Qwen3.8 Max comes out ahead, 83 to 70 and 60 on our weighted score, though DeepSeek V3.2 is 9.5× cheaper per token.
Price in rating: Excluded · Include price
DeepSeek
DeepSeek V3.2
60/100- ECI146.3
- Price$0.28 / $0.42
- Context128K
OpenAI
GPT-5.4 nano
70/100- ECI146.3
- Price$0.20 / $1.25
- Context400K
- Our pick
Alibaba (Qwen)
Qwen3.8 Max
83/100- ECI157.6
- Price$2.00 / $6.00
- Context1M
Qwen3.8 Max is our pick
Qwen3.8 Max is the better all-round choice, scoring 83/100 against GPT-5.4 nano (70) and DeepSeek V3.2 (60). It leads on capability, inputs & features and context window. The score weighs capability 67%, inputs & features 20%, context window 13%.
- CapabilityQwen3.8 MaxCapabilities Index (ECI): Qwen3.8 Max 157.6 · GPT-5.4 nano 146.3 · DeepSeek V3.2 146.3
- Lowest priceDeepSeek V3.2DeepSeek V3.2 $0.315 · GPT-5.4 nano $0.463 · Qwen3.8 Max $3.00 per 1M tokens (3:1 blend)
- Longest contextQwen3.8 MaxQwen3.8 Max 1,000,000 · GPT-5.4 nano 400,000 · DeepSeek V3.2 128,000 tokens
- Widest inputsQwen3.8 MaxDeepSeek V3.2: Text · GPT-5.4 nano: Text, Images · Qwen3.8 Max: Text, Images, PDFs, Video
- Self-hostingDeepSeek V3.2Publishes downloadable weights (MIT License)
| Measure | Weight | DeepSeek V3.2 | GPT-5.4 nano | Qwen3.8 Max |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 67% | 74 | 74 | 84 |
| Inputs & features | 20% | 35 | 70 | 90 |
| Context window | 13% | 32 | 50 | 64 |
| Overall | 100% | 60/100 | 70/100 | 83/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 146.3 | 146.3 | 157.6 (best) |
| ECI rank | #87 of 158 | #86 of 158 | #30 of 158 (best) |
| GPQA DiamondGraduate-level science questions | 83.4% | 78.5% | 92.7% (best) |
| FrontierMath Tiers 1–3Research-level mathematics | — | 44.9% | 74.7% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | 87.8% | 87.8% | 99.4% (best) |
| SimpleQA VerifiedShort factual questions | — | 11.7% | 45.8% (best) |
| Price per million tokens | |||
| Input | $0.28 | $0.20 (best) | $2.00 |
| Output | $0.42 (best) | $1.25 | $6.00 |
| Cached input | — | $0.02 (best) | $0.25 |
| Blended (3:1) | $0.315 (best) | $0.463 | $3.00 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 30 providers | Official OpenAI API | Official Alibaba API |
| Limits | |||
| Context window | 128,000 tokens | 400,000 tokens | 1,000,000 tokens (best) |
| Max output | 64,000 tokens | 128,000 tokens | 131,072 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | Yes |
| Audio | No | No | No |
| Video | No | No | Yes |
| Reasoning | Yes | Yeslow · medium · high · xhigh | Yeslow · medium · xhigh |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | Yes |
| Availability | |||
| Weights | OpenMIT License | Proprietary | Proprietary |
| API model ID | — | gpt-5.4-nano | qwen3.8-max |
| API providers | 30 | 36 | 38 (best) |
| Released | Dec 1, 2025 | Mar 17, 2026 | Aug 3, 2026 |
| Knowledge cutoff | Jul 2024 | Aug 31, 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 V3.2$3.64
GPT-5.4 nano$4.50
Qwen3.8 Max$32.00
Which should you choose?
Which is better: DeepSeek V3.2, GPT-5.4 nano or Qwen3.8 Max?
Qwen3.8 Max is the better all-round choice, scoring 83/100 against GPT-5.4 nano (70) and DeepSeek V3.2 (60). It leads on capability, inputs & features and context window. The score weighs capability 67%, inputs & features 20%, context window 13%.
Which is cheaper, DeepSeek V3.2, GPT-5.4 nano or Qwen3.8 Max?
DeepSeek V3.2 is cheaper at $0.28 input / $0.42 output per million tokens (median across 30 API providers). GPT-5.4 nano costs $0.20 input / $1.25 output per million tokens (official OpenAI API price); Qwen3.8 Max costs $2.00 input / $6.00 output per million tokens (official Alibaba API price). At a typical mix of three input tokens to one output token, that is $0.315 per million tokens for DeepSeek V3.2 versus $0.463 for GPT-5.4 nano (1.5× as much) and $3.00 for Qwen3.8 Max (9.5× as much).
Which scores higher on benchmarks?
Qwen3.8 Max scores higher on the Capabilities Index (ECI): Qwen3.8 Max 157.6 (#30 of 158), GPT-5.4 nano 146.3 (#86 of 158) and DeepSeek V3.2 146.3 (#87 of 158). Their confidence ranges do not overlap (155.5–159.9 vs 144.7–148.0), so the gap is a real one. On individual benchmarks: GPQA Diamond — Qwen3.8 Max 92.7%, DeepSeek V3.2 83.4%, GPT-5.4 nano 78.5%; OTIS Mock AIME 2024–2025 — Qwen3.8 Max 99.4%, DeepSeek V3.2 87.8%, GPT-5.4 nano 87.8%.
Which is better for coding?
There are no published SWE-bench Verified results for DeepSeek V3.2, GPT-5.4 nano and Qwen3.8 Max yet, so there is no like-for-like coding score. On overall capability, Qwen3.8 Max 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?
Qwen3.8 Max has the largest context window at 1,000,000 tokens, against 400,000 for GPT-5.4 nano and 128,000 for DeepSeek V3.2. Maximum output per response: DeepSeek V3.2 up to 64,000, GPT-5.4 nano up to 128,000, Qwen3.8 Max up to 131,072 tokens.
Which can read images, PDFs, audio or video?
DeepSeek V3.2 accepts text; GPT-5.4 nano accepts text and images; Qwen3.8 Max accepts text, images, PDFs and video. Qwen3.8 Max handles the widest range of inputs.
Are any of these open source?
DeepSeek V3.2 publishes its weights (MIT License) and can be self-hosted; GPT-5.4 nano and Qwen3.8 Max is proprietary.
Which is newer?
Qwen3.8 Max is the newest, released Aug 3, 2026. GPT-5.4 nano came out Mar 17, 2026; DeepSeek V3.2 came out Dec 1, 2025. Knowledge cutoff: DeepSeek V3.2 Jul 2024, GPT-5.4 nano Aug 31, 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.