HAIL.SO / DISPATCH · 2026-07-24 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

4 models
Meta
Llama 4 Maverick
×
Alibaba
Qwen3-Max
×
Perplexity
Sonar
×
xAI
Grok 4.5
×
Output $/MTok$0.85$6.00$1.00$6.00
Input $/MTok$0.27$1.20$1.00$2.00
Cached input $/MTok$0.3000
Context window1,048,576262,144128,000500,000
Output cap8,19232,7688,000500,000
Modalitiesin: text, image / out: textin: text / out: textin: text / out: textin: text, image / out: text
Tool use
Structured output
FamilyLlama 4Qwen 3 MaxSonarGrok 4
Knowledge cutoff2024-08-312025-06-302026-02-01
Verified2026-07-192026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗provider page ↗
03

Text to speech

2 models
LMNT
LMNT Blizzard
×
ElevenLabs
Eleven v3
×
$/1M chars$50.00$100.00
Voice qualityneuralneural
Voice cloning✓ included✓ included
Voice count
Languages31+70+
SSML support
TTFB
Output formats
Verified2026-07-192026-07-19
Sourceprovider page ↗provider page ↗