Choose Moondream: Moondream 3 Preview if…
- you want the newer model — released 2026-10-15, 5 months after the other
Moondream: Moondream 3 Preview vs Perceptron: Perceptron Mk1
Neither wins outright: Moondream: Moondream 3 Preview is the pick when you want the newer model, Perceptron: Perceptron Mk1 when cost matters at scale.
Key numbers
| Spec | Moondream: Moondream 3 Preview | Perceptron: Perceptron Mk1 |
|---|---|---|
| Provider | Moondream | Perceptron |
| Input price (per 1M tokens) | $0.30 | $0.15 |
| Output price (per 1M tokens) | $2.50 | $1.50 |
| Blended price (3:1 input/output) | $0.85 | $0.49 |
| Cheaper than … of all models we track | 43% | 54% |
| Context window | 32K tokens (~48 pages) | 33K tokens (~49 pages) |
| Larger context than … of all models | 5% | 6% |
| Max output per response | 4K tokens (~3,100 words) | 8K tokens (~6,100 words) |
| Input types | text, image | text, image, video |
| Output types | text | text |
| Free variant | No | No |
| Listed since | 2026-10-15 | 2026-05-12 |
What it costs for real workloads
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) | Moondream: Moondream 3 Preview | Perceptron: Perceptron Mk1 | Difference |
|---|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation (1,500 in / 400 out tokens) | $72.50 | $41.25 | Perceptron: Perceptron Mk1 saves $31.25 |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages (6,000 in / 500 out tokens) | $61.00 | $33.00 | Perceptron: Perceptron Mk1 saves $28.00 |
| 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) | Doesn't fit context | Doesn't fit context | — |
| Batch summarisation 10,000 documents — each a 6,000-word report summarised to a page (8,000 in / 600 out tokens) | $39.00 | $21.00 | Perceptron: Perceptron Mk1 saves $18.00 |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass (100,000 in / 2,000 out tokens) | Doesn't fit context | Doesn't fit context | — |
Capabilities side by side
| Capability | Moondream: Moondream 3 Preview | Perceptron: Perceptron Mk1 | What it lets you do |
|---|---|---|---|
| Tool / function calling | — | — | call your APIs and run agent loops |
| Structured output (JSON schema) | — | ✓ 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) | — | — | repeat a generation for tests and evals |
| Log-probabilities | — | — | read token confidence for classification and scoring |
From each model's published parameters and input/output types. “—” means not advertised, not proven absent.
Price history
- Moondream: Moondream 3 Preview: price unchanged since we started tracking it on 2026-10-16.
- Perceptron: Perceptron Mk1: price unchanged since we started tracking it on 2026-08-06.
Which should you choose?
Choose Perceptron: Perceptron Mk1 if…
- cost matters at scale — it is 43% cheaper on a typical 3:1 input/output mix ($41.25 vs $72.50 a month for 50,000 support chats)
- you need structured output (JSON schema) — to return JSON that validates against your schema; Moondream: Moondream 3 Preview does not advertise it
- you generate long outputs in one go — up to 8K tokens (about 6,100 words) per response
Frequently asked questions
Is Moondream: Moondream 3 Preview better than Perceptron: Perceptron Mk1?
Neither wins outright: Moondream: Moondream 3 Preview is the pick when you want the newer model, Perceptron: Perceptron Mk1 when cost matters at scale. In short: Moondream: Moondream 3 Preview when you want the newer model; Perceptron: Perceptron Mk1 when cost matters at scale, you need structured output (JSON schema), you generate long outputs in one go.
Which is cheaper, Moondream: Moondream 3 Preview or Perceptron: Perceptron Mk1?
Perceptron: Perceptron Mk1 is cheaper: $0.49 vs $0.85 per 1M tokens at a typical 3:1 input/output mix, 43% less. For 50,000 support conversations a month that is $41.25 against $72.50.
How much do Moondream: Moondream 3 Preview and Perceptron: Perceptron Mk1 cost per million tokens?
Moondream: Moondream 3 Preview costs $0.30 per 1M input tokens and $2.50 per 1M output tokens; Perceptron: Perceptron Mk1 costs $0.15 input and $1.50 output. Output tokens are billed separately, so long answers shift the gap.
Which has the bigger context window, Moondream: Moondream 3 Preview or Perceptron: Perceptron Mk1?
Perceptron: Perceptron Mk1: 33K tokens, about 49 pages, against 32K (about 48 pages) for Moondream: Moondream 3 Preview.
Do Moondream: Moondream 3 Preview and Perceptron: Perceptron Mk1 support function calling and JSON output?
Tool calling — Moondream: Moondream 3 Preview: not advertised, Perceptron: Perceptron Mk1: not advertised. JSON-schema structured output — Moondream: Moondream 3 Preview: not advertised, Perceptron: Perceptron Mk1: yes.
Can Moondream: Moondream 3 Preview or Perceptron: Perceptron Mk1 read images?
Yes, both accept image input.
Is Moondream: Moondream 3 Preview or Perceptron: Perceptron Mk1 free to use?
No free variant is listed for either; both are pay-per-token.
What is the maximum output length of Moondream: Moondream 3 Preview and Perceptron: Perceptron Mk1?
Moondream: Moondream 3 Preview can write up to 4K tokens (about 3,100 words) per response; Perceptron: Perceptron Mk1 up to 8K (about 6,100 words).
When were Moondream: Moondream 3 Preview and Perceptron: Perceptron Mk1 released?
Moondream: Moondream 3 Preview was listed on 2026-10-15; Perceptron: Perceptron Mk1 on 2026-05-12.
About the two models
Moondream: Moondream 3 Preview
Moondream 3 Preview is a vision-language model from Moondream with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and visual reasoning over images, with...
Full Moondream: Moondream 3 Preview specs →Perceptron: Perceptron Mk1
Perceptron Mk1 (Mark One) is Perceptron's highest-quality vision-language model for video and embodied reasoning.** It accepts image and video inputs paired with natural language queries, and produces detailed visual understanding...
Full Perceptron: Perceptron Mk1 specs →