HAIL.SO / DISPATCH · 2026-07-25 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

5 models
Perplexity
Sonar Pro
×
OpenAI
OpenAI o1
×
Perplexity
Sonar
×
DeepSeek
DeepSeek V3
×
Alibaba
Qwen3-Max
×
Output $/MTok$15.00$60.00$1.00$0.28$6.00
Input $/MTok$3.00$15.00$1.00$0.14$1.20
Cached input $/MTok$7.5000$0.0028
Context window200,000200,000128,0001,048,576262,144
Output cap8,000100,0008,000384,00032,768
Modalitiesin: text / out: textin: text, image / out: textin: text / out: textin: text / out: textin: text / out: text
Tool use
Structured output
FamilySonaro-seriesSonarDeepSeek V3Qwen 3 Max
Knowledge cutoff2023-10-012025-06-30
Verified2026-07-192026-07-192026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗provider page ↗provider page ↗
03

Text to speech

1 model
Smallest.ai
Lightning v3.1 Pro
×
$/1M chars$19.50
Voice qualityneural
Voice cloning
Voice count
Languagesen, hi, mr, ta, ml, te, kn, pa, bn, or, gu, ar, zh, id, ja, ko, ms, tr, vi, de, es, fr, it, pt, ru, el, fi, no, pl
SSML support
TTFB200ms
Output formatspcm, mp3, wav, ulaw, alaw
Verified2026-07-19
Sourceprovider page ↗