Meta: Llama 4 Scout
Decision snapshot from the global rankings; composite scoring matches the on-site model board.
Data updated:
Model
1 verified sources
Aggregator quote only — no official sync yet
Next steps on AI Hippo
About this model
Llama 4 Scout is Meta's efficiency-oriented Llama 4 variant in this snapshot—positioned for strong context and throughput at a lower price tier than larger siblings. Use the metrics dash, price tables, Compare, and Token hub to decide whether it fits your workload.
You can also explore more models from Meta , and browse more options from 🇺🇸 United States .
Capabilities & specs
- Modality
- text+image->text
- Input modalities
- text, image
- Knowledge cutoff
- 2024-08-31
- Supported API parameters
- frequency_penalty, logit_bias, max_tokens, min_p, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
Catalog description: Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input...
Token pricing by provider
Compare per-provider token prices for this model across available platforms.
| Provider | Input / 1M tokens | Output / 1M tokens | Latency | Status |
|---|---|---|---|---|
| OpenRouter | $0.10 | $0.30 | — | Verified · 2026-08-02 |
Provider prices are sourced from the token comparison dataset and may change between snapshots.
Alternative picks
-
Xiaomi: MiMo-V2.5
Compare now -
OpenAI: GPT-5.6 Luna Pro
Compare now -
OpenAI GPT Latest
Compare now
Pick one or two more models on global rankings and use Compare to view them side by side.
FAQ
Why is the official price missing?
Official rows require a verified direct vendor or cloud API price. When only aggregators list a model, the Official column shows — and verified aggregator quotes appear in the token table.
How is context different from max output tokens?
Context window is how much input the model can accept in a single request. Max output tokens is the per-response generation cap from the primary listing—often much smaller than the context window.
Where do multi-provider prices come from?
Prices are crawled or synced from token providers (OpenRouter, Groq, Together, and others) and merged into the infrastructure comparison dataset. Verified rows show a fetch date in the status column.
Data source & methodology
- Source
- OpenRouter catalog, verified vendor crawls, and Hugging Face hub snapshots.
- Metrics
- Context window, blended 1M-token price, max output tokens, and capability fields from the OpenRouter models API.
- Update cadence
- Daily pipeline refresh; provider prices may change between snapshots.
- Fetched at
- 2026-08-02
- Method
- Composite rank uses on-site context × price weighting (not LMSYS Arena ELO).