open-source

Llama 4 Maverick 17b 128e Instruct vs Claude Sonnet 4 5

Side-by-side pricing, context window, features, and per-use-case verdict. Data sourced from LiteLLM, updated regularly.

Specifications

Feature Llama 4 Maverick 17b 128e Instruct Claude Sonnet 4 5
Provider groq anthropic
Model ID groq/meta-llama/llama-4-maverick-17b-128e-instruct claude-sonnet-4-5
Input / 1M tokens $0.20 (93% cheaper) $3.00
Output / 1M tokens $0.60 (96% cheaper) $15.00
Cached input / 1M $0.30
Context window 131kk tokens 200kk tokens
Max output 8kk tokens 64kk tokens
Vision
Function calling
Prompt caching
Web search
JSON schema

Monthly cost comparison

No cache discount applied — pure token costs at three usage scales.

Scale Llama 4 Maverick 17b 128e Instruct Claude Sonnet 4 5 Winner
Small
1M in / 0.2M out
$0.32 $6.00 Llama 4 Maverick 17b 128e Instruct saves 95%
Medium
10M in / 2M out
$3.20 $60.00 Llama 4 Maverick 17b 128e Instruct saves 95%
Large
100M in / 20M out
$32.00 $600 Llama 4 Maverick 17b 128e Instruct saves 95%

Category winners

Cheapest input
Llama 4 Maverick 17b 128e Instruct
$0.20/M · 93% less than the other
Cheapest output
Llama 4 Maverick 17b 128e Instruct
$0.60/M · 96% less than the other
Larger context
Claude Sonnet 4 5
200kk tokens · 53% larger window
More features
Claude Sonnet 4 5
4 / 5 capabilities

Better for…

High-volume RAG & repeated context
$0.30/M cached input — prompt caching makes repeated context retrieval significantly cheaper
Claude Sonnet 4 5
Short interactions & high-throughput APIs
$0.60/M output — 96% cheaper on output at scale
Llama 4 Maverick 17b 128e Instruct
Long-document processing & large codebases
200k context window vs 131k — 53% larger context
Claude Sonnet 4 5
Agentic workflows & tool use
Supports 4 of 5 tracked capabilities (vision, function calling, caching, web search, JSON schema)
Claude Sonnet 4 5

Deep-dive on each model

Pricing data sourced from LiteLLM and refreshed regularly. Always verify with each provider's official pricing page before making business decisions.