WWhichAITry via Vercel AI Gateway
Models/Meta
Meta · released 2025-07open

Llama 4 405B

Open-weights heavyweight. Self-host or run anywhere.

context
256K
output max
16K
input /M
$2.70
output /M
$2.70
modalities
T·I
Verdict

What it's actually good at

Strengths

  • +Open weights — self-host
  • +No vendor lock-in
  • +Flat input/output pricing

Weaknesses

  • Requires infra or paid inference host

Our scores

reasoning
8.0
coding
8.0
writing
8.0
speed
6.0
value
8.0
Meta's open-weights flagship. Competitive with proprietary models while giving you full deployment control.
Pricing

Cost at common volumes

Monthly volumeInputOutputEst. monthly
Side project1M tok0.3M tok$3.51
Growing app50M tok15M tok$175.50
Production500M tok150M tok$1755.00
Scale5000M tok1500M tok$17550.00

Estimates assume uncached input. Prompt caching and batch APIs can cut this by 50–90% for many workloads. Use the calculator →

Head-to-head

Compare Llama 4 405B