OpenRouter model
Meta: 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...
- Provider
- OpenRouter
- Availability
- Available
- Model ID
- meta-llama/llama-4-maverick
- Snapshot
- 2026-10-07
Meta: Llama 4 Maverick Benchmarks & Agentic Accuracy
Verified primary research & developer preference metrics
What This Means in Real Software Development
- 61.2% SWE-bench Verified score marks a major milestone for open-weights autonomous software engineering.
- 78.5% on AIME confirms elite multi-step deduction and algorithmic problem solving.
- Native multimodal vision allows inspecting web UI screenshots and architecture diagrams directly.
- Available across open API providers and for private enterprise GPU cluster hosting.
Swarm Multi-Agent Topology
Best Paired With: Claude 3.7 Sonnet as Lead Planner
High-fidelity code generation and vision capabilities make it ideal for full-stack web and systems coding.
Executes isolated feature work inside dedicated Git worktrees.
What Developers & Practitioners Are Saying
Real-world field observations, production strengths, and developer consensus
"Next-generation open architecture from Meta, featuring enhanced multimodal understanding and modernized code pretraining."
Verified Real-World Strengths
- 61.2% on SWE-bench Verified
- Strong native vision capabilities for web UI generation and wireframe translation
- Fast inference on modern tensor core architectures
Production Caveats & Gaps
- Newer architecture with evolving quantization support across edge engines
Primary Sources & Methodology Citations
- Meta Llama 4 Technical Preview — Meta AI (March 2026)
Head-to-Head Frontier Comparisons: Meta: Llama 4 Maverick vs Rivals
Standardized, side-by-side benchmark & economics comparison against peer models
| Model | Role / Tier | SWE-bench | LiveCodeBench | LMSYS Elo | AIME | Input / Output |
|---|---|---|---|---|---|---|
Meta: Llama 4 MaverickOpenRouterCurrent | Tier 2: Next-Gen Open Frontier Engine | 61.2% | 55.4% | 1370 | 78.5% | $0.8 / $2.4 |
Mistral Devstral 2512Mistral Compare Meta: Llama 4 Maverick vs Mistral Devstral 2512 → | Tier 2: Dedicated Autonomous Coder | 59.2% | 54% | 1360 | 71.5% | $0.5 / $1.5 |
OpenAI o4-miniOpenAI Compare Meta: Llama 4 Maverick vs OpenAI o4-mini → | Tier 2: High-Speed Deductive Engine | 56.8% | 73.4% | 1385 | 90.1% | $1.2 / $4.8 |
DeepSeek R1DeepSeek Compare Meta: Llama 4 Maverick vs DeepSeek R1 → | Tier 2: Deep Algorithmic Specialist | 49.2% | 49.8% | 1360 | 79.8% | $0.55 / $2.19 |
Llama 4 ScoutMeta Compare Meta: Llama 4 Maverick vs Llama 4 Scout → | Tier 2: Global Context Specialist | 69% | 61% | 1380 | — | $0.6 / $2.4 |
Meta: Llama 4 Maverick model overview
Meta: Llama 4 Maverick is listed in the public Swarm catalog through OpenRouter. This page summarizes the verified catalog fields used by Swarm without claiming real-time access for every provider account.
- Provider
- OpenRouter
- Family
- OpenRouter
- Context window
- 1.05M tokens
- Thinking support
- Not supported
- Availability status
- Available
- Catalog ID
- openrouter/meta-llama/llama-4-maverick
Run Meta: Llama 4 Maverick in Swarm
Verified CLI workflow & multi-agent architecture
swarm ctl model set --provider openrouter --model openrouter/meta-llama/llama-4-maverick# Or switch interactively inside the terminal TUI:
/modelsSubagent Role Matrix
Assign Meta: Llama 4 Maverick to specialized system roles in ~/.swarm/config.json:
- Primary / Planner: Orchestrates full task DAGs and multi-stage pipelines.
- Coder Subagent: Generates code and unit tests inside private Git worktrees.
- Finder Subagent: Performs read-only code search, symbol discovery, and mapping.
Public catalog snapshot. Catalog availability is not a real-time provider status or an access guarantee.