Model specification · Meta
Llama 4 Maverick
meta-llama/llama-4-maverick
Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...
SpecificationsReported values
- Context window
- 1.0M
- Maximum output
- 16K
- Architecture
- text+image->text
- Tokenizer
- Llama4
- Knowledge cutoff
- 2024-08-31
- Moderated
- No
- Input modalities
- text image
- Output modalities
- text
Full pricingUSD where reported
| Price | Standard |
|---|---|
| Input | $0.200 / M |
| Output | $0.800 / M |
Benchmark indexesReported indexes
- Intelligence
- 14.3
- Coding
- 16.3
- Agentic
- 1.3
- Design Arena Elo
- 916
Example workload costsStandard pricing only
| Workload | Input / output | Cache share | Estimated cost |
|---|---|---|---|
| Quick check | 10K / 1K × 1 | 0% | $0.00280 |
| Long analysis | 100K / 4K × 1 | 50% | $0.023 |
| Workday | 20K / 2K × 1,000 | 30% | $5.60 |
More from MetaAll Meta models
| Model | Context | Input / M | Output / M | Intel. |
|---|---|---|---|---|
| Llama 4 Scout | 1.3M | $0.100 | $0.300 | 10 |
| Llama 3.3 70B Instruct | 131K | $0.130 | $0.400 | 9.4 |
| Llama 3.1 8B Instruct | 131K | $0.050 | $0.080 | 7.6 |