Neither wins outright: DeepSeek: DeepSeek V4 Flash Vision Exp (batch) is the pick when cost matters at scale, NVIDIA: Nemotron 3.5 Super VL when you need seed (reproducible sampling).
| Spec | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) | NVIDIA: Nemotron 3.5 Super VL |
|---|---|---|
| Provider | DeepSeek | Nvidia |
| Input price (per 1M tokens) | $0.11 | $0.15 |
| Output price (per 1M tokens) | $0.33 | $0.50 |
| Blended price (3:1 input/output) | $0.17 | $0.24 |
| Cheaper than … of all models we track | 73% | 67% |
| Context window | 1.0M tokens (~1,573 pages) | 1.0M tokens (~1,573 pages) |
| Larger context than … of all models | 81% | 81% |
| Max output per response | 944K tokens (~707,800 words) | 236K tokens (~176,900 words) |
| Input types | text, image | text, image, video |
| Output types | text | text |
| Free variant | No | Yes |
| Listed since | 2026-08-21 | 2026-10-15 |
Monthly cost at list price for five common workloads. Token counts per unit are stated so you can scale them to your own traffic. A workload that does not fit a model's context window is marked rather than priced.
| Workload (per month) | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) | NVIDIA: Nemotron 3.5 Super VL | Difference |
|---|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation (1,500 in / 400 out tokens) | $14.85 | $21.25 | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) saves $6.40 |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages (6,000 in / 500 out tokens) | $16.50 | $23.00 | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) saves $6.50 |
| Coding / tool-using agent 2,000 tasks — each a multi-step task with tool calls and file context (60,000 in / 6,000 out tokens) | $17.16 | $24.00 | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) saves $6.84 |
| Batch summarisation 10,000 documents — each a 6,000-word report summarised to a page (8,000 in / 600 out tokens) | $10.78 | $15.00 | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) saves $4.22 |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass (100,000 in / 2,000 out tokens) | $5.83 | $8.00 | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) saves $2.17 |
| Capability | DeepSeek: DeepSeek V4 Flash Vision Exp (batch) | NVIDIA: Nemotron 3.5 Super VL | What it lets you do |
|---|---|---|---|
| Tool / function calling | ✓ Yes | ✓ Yes | call your APIs and run agent loops |
| Structured output (JSON schema) | ✓ Yes | ✓ Yes | return JSON that validates against your schema |
| Reasoning / thinking mode | ✓ Yes | ✓ Yes | spend extra tokens thinking before answering hard problems |
| Image input | ✓ Yes | ✓ Yes | read screenshots, charts and scanned pages |
| Audio input | — | — | take speech or audio directly |
| File / PDF input | — | — | accept documents without your own parsing step |
| Image output | — | — | generate images, not just text |
| Built-in web search | — | — | answer from live web results |
| Seed (reproducible sampling) | — | ✓ Yes | repeat a generation for tests and evals |
| Log-probabilities | ✓ Yes | ✓ Yes | read token confidence for classification and scoring |
From each model's published parameters and input/output types. “—” means not advertised, not proven absent.
Neither wins outright: DeepSeek: DeepSeek V4 Flash Vision Exp (batch) is the pick when cost matters at scale, NVIDIA: Nemotron 3.5 Super VL when you need seed (reproducible sampling). In short: DeepSeek: DeepSeek V4 Flash Vision Exp (batch) when cost matters at scale, you generate long outputs in one go; NVIDIA: Nemotron 3.5 Super VL when you need seed (reproducible sampling), you want to prototype at zero cost.
DeepSeek: DeepSeek V4 Flash Vision Exp (batch) is cheaper: $0.17 vs $0.24 per 1M tokens at a typical 3:1 input/output mix, 31% less. For 50,000 support conversations a month that is $14.85 against $21.25.
DeepSeek: DeepSeek V4 Flash Vision Exp (batch) costs $0.11 per 1M input tokens and $0.33 per 1M output tokens; NVIDIA: Nemotron 3.5 Super VL costs $0.15 input and $0.50 output. Output tokens are billed separately, so long answers shift the gap.
Both take 1.0M tokens per request — roughly 1,573 pages of text.
Tool calling — DeepSeek: DeepSeek V4 Flash Vision Exp (batch): yes, NVIDIA: Nemotron 3.5 Super VL: yes. JSON-schema structured output — DeepSeek: DeepSeek V4 Flash Vision Exp (batch): yes, NVIDIA: Nemotron 3.5 Super VL: yes.
Yes, both accept image input.
NVIDIA: Nemotron 3.5 Super VL has a free variant (nvidia/nemotron-3.5-super-vl:free). Free variants usually carry tighter rate limits.
DeepSeek: DeepSeek V4 Flash Vision Exp (batch) can write up to 944K tokens (about 707,800 words) per response; NVIDIA: Nemotron 3.5 Super VL up to 236K (about 176,900 words).
DeepSeek: DeepSeek V4 Flash Vision Exp (batch) was listed on 2026-08-21; NVIDIA: Nemotron 3.5 Super VL on 2026-10-15.
DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of [DeepSeek V4 Flash 0731](https://openrouter.ai/deepseek/deepseek-v4-flash-0731) from DeepSeek, adding image understanding while matching the base model on text capabilities including agents,...
Full DeepSeek: DeepSeek V4 Flash Vision Exp (batch) specs →NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video...
Full NVIDIA: Nemotron 3.5 Super VL specs →