HAIL.SO / DISPATCH · 2026-06-29 · COMPARE
FILE: COMPARE · 6 of 6 slots

side by side.

← all costsclear
01

Large language models

1 model
DeepSeek
DeepSeek V3
×
Output $/MTok$0.28
Input $/MTok$0.14
Cached input $/MTok$0.0028
Context window1,048,576
Output cap384,000
Modalitiesin: text / out: text
Tool use
Structured output
FamilyDeepSeek V3
Knowledge cutoff
Verified2026-05-18stale 42d
Sourceprovider page ↗
02

Speech to text

3 models
AssemblyAI
Universal-Streaming
×
AssemblyAI
Universal-2
×
Speechmatics
Speechmatics Enhanced
×
$/min$0.002500$0.002500$0.004000
$/min batch
Streaming
Realtime
Languagesen99+55+
Diarizationextra-costextra-costincluded
TTFW
Verified2026-05-05stale 55d2026-05-05stale 55d2026-05-19stale 41d
Sourceprovider page ↗provider page ↗provider page ↗
03

Text to speech

2 models
Smallest.ai
Lightning v3.1
×
Deepgram
Deepgram Aura 2
×
$/1M chars$25.00$30.00
Voice qualityneuralneural
Voice cloning✓ included
Voice count217
Languagesen, hi, es, mr, kn, ta, bn, gu, te, ml, pa, oren, es, de, fr, nl, it, ja
SSML support
TTFB200ms
Output formatspcm, wav, mp3, mulawwav, mp3, linear16, mulaw, alaw, opus
Verified2026-05-19stale 41d2026-05-19stale 41d
Sourceprovider page ↗provider page ↗