HAIL.SO / DISPATCH · 2026-07-23 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

3 models
Alibaba
Qwen3.7-Plus
×
OpenAI
OpenAI o3
×
Mistral
Mistral Large 3
×
Output $/MTok$1.60$8.00$1.50
Input $/MTok$0.40$2.00$0.50
Cached input $/MTok$0.5000
Context window1,048,576200,000262,144
Output cap65,536100,0008,192
Modalitiesin: text, image / out: textin: text, image / out: textin: text, image / out: text
Tool use
Structured output
FamilyQwen 3.7 Pluso-seriesMistral Large
Knowledge cutoff2024-06-01
Verified2026-07-192026-07-192026-07-19
Sourceprovider page ↗provider page ↗provider page ↗
02

Speech to text

2 models
Rev.ai
Rev.ai Reverb
×
Speechmatics
Speechmatics Enhanced
×
$/min$0.003300$0.002150
$/min batch
Streaming
Realtime
Languagesen56+
Diarizationincludedincluded
TTFW
Verified2026-07-192026-07-19
Sourceprovider 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 ↗