Compare LLM Models
Side-by-side comparison of 1 model. Best values highlighted.
| Model ID | meta-llama/Llama-4-Maverick-17B-128E |
| Status | ga |
| Knowledge cutoff | March 2025 |
| Context window | 1M tokens |
| Max output | 8K tokens |
| Parameters | 400B (17B active) |
| Open weights | Yes |
| Pricing (per MTok) | |
| Input | $0.15 |
| Output | $0.60 |
| Cache read | — |
| Cache write | — |
| Free tier | No |
| Capabilities | |
| Tool Use | |
| Streaming | |
| Prompt Caching | |
| Batch API | |
| Extended Thinking | |
| Structured Outputs | |
| Multi-turn Tool Calling | |
| Agentic Workload Ready | |
| Parallel Tool Calls | |
| Vision Input | |
| Audio Input | |
| Audio Output | |
| Modalities | |
| Input | text, vision, code |
| Output | text, code |
| Last updated | May 9, 2026 |