Neither wins outright: Baidu: ERNIE 4.5 VL 424B A47B is the pick when cost matters at scale, Moondream: Moondream 3 Preview when you want the newer model.
| Spec | Baidu: ERNIE 4.5 VL 424B A47B | Moondream: Moondream 3 Preview |
|---|---|---|
| Provider | Baidu | Moondream |
| Input price (per 1M tokens) | $0.42 | $0.30 |
| Output price (per 1M tokens) | $1.25 | $2.50 |
| Blended price (3:1 input/output) | $0.63 | $0.85 |
| Cheaper than … of all models we track | 49% | 43% |
| Context window | 123K tokens (~185 pages) | 32K tokens (~48 pages) |
| Larger context than … of all models | 14% | 5% |
| Max output per response | 16K tokens (~12,000 words) | 4K tokens (~3,100 words) |
| Input types | image, text | text, image |
| Output types | text | text |
| Free variant | No | No |
| Listed since | 2025-06-30 | 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) | Baidu: ERNIE 4.5 VL 424B A47B | Moondream: Moondream 3 Preview | Difference |
|---|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation (1,500 in / 400 out tokens) | $56.50 | $72.50 | Baidu: ERNIE 4.5 VL 424B A47B saves $16.00 |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages (6,000 in / 500 out tokens) | $62.90 | $61.00 | Moondream: Moondream 3 Preview saves $1.90 |
| 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) | $65.40 | 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) | $41.10 | $39.00 | Moondream: Moondream 3 Preview saves $2.10 |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass (100,000 in / 2,000 out tokens) | $22.25 | Doesn't fit context | — |
| Capability | Baidu: ERNIE 4.5 VL 424B A47B | Moondream: Moondream 3 Preview | What it lets you do |
|---|---|---|---|
| Tool / function calling | — | — | call your APIs and run agent loops |
| Structured output (JSON schema) | — | — | 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 | — | — | 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: Baidu: ERNIE 4.5 VL 424B A47B is the pick when cost matters at scale, Moondream: Moondream 3 Preview when you want the newer model. In short: Baidu: ERNIE 4.5 VL 424B A47B when cost matters at scale, you need to fit long inputs in one request, you need seed (reproducible sampling), you generate long outputs in one go; Moondream: Moondream 3 Preview when you want the newer model.
Baidu: ERNIE 4.5 VL 424B A47B is cheaper: $0.63 vs $0.85 per 1M tokens at a typical 3:1 input/output mix, 26% less. For 50,000 support conversations a month that is $56.50 against $72.50.
Baidu: ERNIE 4.5 VL 424B A47B costs $0.42 per 1M input tokens and $1.25 per 1M output tokens; Moondream: Moondream 3 Preview costs $0.30 input and $2.50 output. Output tokens are billed separately, so long answers shift the gap.
Baidu: ERNIE 4.5 VL 424B A47B: 123K tokens, about 185 pages, against 32K (about 48 pages) for Moondream: Moondream 3 Preview.
Tool calling — Baidu: ERNIE 4.5 VL 424B A47B: not advertised, Moondream: Moondream 3 Preview: not advertised. JSON-schema structured output — Baidu: ERNIE 4.5 VL 424B A47B: not advertised, Moondream: Moondream 3 Preview: not advertised.
Yes, both accept image input.
No free variant is listed for either; both are pay-per-token.
Baidu: ERNIE 4.5 VL 424B A47B can write up to 16K tokens (about 12,000 words) per response; Moondream: Moondream 3 Preview up to 4K (about 3,100 words).
Baidu: ERNIE 4.5 VL 424B A47B was listed on 2025-06-30; Moondream: Moondream 3 Preview on 2026-10-15.
ERNIE-4.5-VL-424B-A47B is a multimodal Mixture-of-Experts (MoE) model from Baidu’s ERNIE 4.5 series, featuring 424B total parameters with 47B active per token. It is trained jointly on text and image data...
Full Baidu: ERNIE 4.5 VL 424B A47B specs →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 →