Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
Scroll to compare models →
| 3 models | Unknowncommand-aRemove × | bytedance-seedByteDance Seed: Seedream 5.0 ProRemove × | moondreamMoondream: Moondream 3 PreviewRemove × |
|---|---|---|---|
| What’s different | 256K context · 8.2K max output · Tool calling · Structured output | Choose standard or fast prompt enhancement; accepts extreme aspect ratios up to 16:1 within a 2K-square pixel budget. | 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… |
| Date | Oct 16, 2026Discovered | Jul 8, 2026Released | Oct 15, 2026Listed |
API prices USD · provider configuration shown below | |||
| Listed rate | $2.50 in / $10.00 out/ 1M tokensExperiential · all routes ↗Standard list rate | $0.0675/ imagefal · all routes ↗Area ≤ 1536x1536 pixels | $0.30 in / $2.50 out/ 1M tokensPrimeIntellect · via OpenRouter · all routes ↗primeintellect/fp8 · quantization=fp8 |
| Provider endpoints | 1 endpoint → | 2 endpoints → | 2 endpoints → |
Benchmarks Each result retains its evaluation configuration | |||
| AA Image ↗2.1 | Not reported | 1,100.01 ↗Text-to-image · all categoriesIndependent result95% CI 1,089.01–1,111.01 · n=11,391 | Not reported |
Specifications | |||
| Context window | 256K | — | 32K |
| Max output | 8.2K | — | 4.1K |
| Parameters | — | — | — |
| Tool calling | Supported | Not reported | Not reported |
| Structured output | Supported | Not reported | Not reported |
| Image input | Not reported | Supported | Supported |
| Reasoning | Not reported | Not reported | Supported |
| Open weights | Not reported | Not reported | Supported |